FIELD

[0001] The embodiments described herein relate to a method and apparatus for the detection of metals and, in particular, to the detection of gold and alloys of gold and silver, commonly called electrum, in mineral and material samples taken in the field.

INTRODUCTION

[0002] Gold can be found naturally in many, if not most, countries of the world. In each case, the concentration and occurrence of gold may vary. In particular, gold may occur in one or more of the following forms: placer gold (dust to nugget size); visible gold; particulate and micron-sized gold (not visible to the eye at mineral grain boundaries); micron-sized gold (particles greater than 0.2 μm across) within host minerals; colloidal gold (particles less than 0.2 pm across); in solid solution (e.g., where gold is substituted for another metal atom in the structure of a mineral); and in gold minerals (e.g., where gold forms chemical bonds with other atoms).

[0003] The detection of gold in particulate, micron and submicron sizes is of particular interest in field exploration and mining applications. The majority of gold deposits typically have gold grains at micron or larger sizes.

[0004] Typical gold grades encountered during field exploration are in the range of 10-1000 parts per billion (ppb). Mining operations typically encounter gold in concentrations of 0.5 to 30 grams per tonne (g/tonne), which may also be expressed as parts per million (ppm).

[0005] One technique that is used in field exploration and mining involves the use of X-ray fluorescence (XRF) to detect the presence of gold, or elements that may indicate the presence of gold such as arsenic. However, portable XRF analyzers are not currently sensitive enough to detect gold at typical economic grades, or at lower anomalous values important for exploration.

SUMMARY

[0006] In a first broad aspect, there is provided a method for detecting gold in a material sample. The method may comprise: illuminating the material sample with light in a first preselected wavelength range and with light in a second preselected wavelength range; detecting the light in the first preselected wavelength range reflected by the material sample and the light in the second preselected wavelength range reflected by the material sample; and comparing, using a processor, a relative intensity of the reflected light in the second preselected wavelength range to the reflected light in the first preselected wavelength range.

[0007] In some embodiments, the material sample can be substantially simultaneously illuminated in the first preselected wavelength range and in the second preselected wavelength range. In some other embodiments, the material sample can be sequentially illuminated in the first preselected wavelength range and in the second preselected wavelength range.

[0008] In some embodiments, b) can be carried out by an image sensor comprising a two-dimensional matrix of sensor elements, and c) may comprise comparing the relative intensity for each sensor element in the two-dimensional matrix of sensor elements to compute a weighted intensity value for each sensor element.

[0009] In some embodiments, the method may further comprise comparing the relative intensity of the reflected light in the second preselected wavelength range detected at each sensor element in the two-dimensional matrix of sensor elements to an average relative intensity of the reflected light in the second preselected wavelength range detected by the two-dimensional matrix of sensor elements.

[0010] In some embodiments, the first preselected wavelength range can be in the BV (blue-violet) spectrum, and the second preselected wavelength range can be in the ROY (red-orange-yellow) spectrum.

[0011] In some embodiments, the reflected light in the first preselected wavelength range may comprise substantially monochromatic light with a wavelength less than 500 nm, and the reflected light in the second preselected wavelength range may comprise substantially monochromatic light with a wavelength greater than 500 nm.

[0012] In some embodiments, the reflected light can be filtered to isolate the first preselected wavelength range, and the reflected light can be separately filtered to isolate the second preselected wavelength range.

[0013] In some embodiments, the method may further comprise detecting ambient light reflected by the material sample, wherein the detected ambient light reflected by the material sample can be used to reduce noise when computing the relative intensity in c). [0014] In some embodiments, the method may further comprise: illuminating the material sample with light in a third preselected wavelength range; and detecting the light reflected by the material sample in the third preselected wavelength range, wherein the comparing further comprises comparing intensity of the reflected light in the third preselected wavelength range to intensity of the reflected light in the first and second preselected wavelength ranges.

[0015] In some embodiments, the first preselected wavelength is in the blue-violet spectrum, the second preselected wavelength is in the red-orange-yellow spectrum, and the third preselected wavelength is in the green spectrum.

[0016] In some embodiments, the method may further comprise identifying a reflected light intensity in the second preselected wavelength range greater than 55% of the maximum intensity.

[0017] In some embodiments, the comparing comprises computing a reflectance slope product, wherein the reflectance slope product is a result of the square of reflected light intensity in the second wavelength multiplied by reflected light intensity in the third wavelength, and divided by the square of reflected light intensity in the first wavelength.

[0018] In another broad aspect, there is provided a method for detecting gold in a material sample, the method comprising: illuminating the material sample with light in a first preselected wavelength range less than 500 nm and with light in a second preselected wavelength range greater than 500 nm; detecting the light in the first preselected wavelength range reflected by the material sample and the light in the second preselected wavelength range reflected by the material sample, wherein the light is detected by an image sensor comprising a two-dimensional matrix of sensor elements; and computing, using a processor, a reflectance ratio product for each individual sensor element in the two-dimensional matrix of sensor elements, wherein the reflectance ratio product comprises the product of a wavelength reflectance ratio and a spatial reflectance ratio, wherein the wavelength reflectance ratio comprises relative intensity of the reflected light in the second preselected wavelength range to the reflected light in the first preselected wavelength range, and wherein the spatial reflectance ratio is computed based on the ratio of relative intensity of the reflected light in the second preselected wavelength range in each individual sensor element to relative intensity of the reflected light in the second preselected wavelength range in a plurality of sensor elements. [0019] In yet another broad aspect, there is provided a method for detecting gold in a material sample, the method comprising: illuminating the material sample with light in a first preselected wavelength range, with light in a second preselected wavelength range and with light in a third preselected wavelength range, wherein the first preselected wavelength is in the blue-violet spectrum, wherein the second preselected wavelength is in the red-orange-yellow spectrum, and wherein the third preselected wavelength is in the green spectrum; detecting the light in the first, second and third preselected wavelength ranges reflected by the material sample, wherein the light is detected by an image sensor comprising a two-dimensional matrix of sensor elements; and computing, using a processor, a reflectance slope product, wherein the reflectance slope product is a result of the square of reflected light intensity in the second wavelength multiplied by reflected light intensity in the third wavelength, and divided by the square of reflected light intensity in the first wavelength.

[0020] In a further broad aspect, there is provided an apparatus for detecting gold in a material sample. The apparatus may comprise: at least one light source for illuminating the material sample with light in a first preselected wavelength range and with light in a second preselected wavelength range; an image sensor for detecting the light in the first preselected wavelength range reflected by the material sample and the light in the second preselected wavelength range reflected by the material sample; and a processor, for comparing a relative intensity of the reflected light in the second preselected wavelength range to the reflected light in the first preselected wavelength range.

[0021] In some embodiments, the image sensor may comprise a two-dimensional matrix of sensor elements, and the processor may compare the relative intensity for each sensor element in the two-dimensional matrix of sensor elements to compute a weighted intensity value for each sensor element.

[0022] In some embodiments, the processor may also be configured to compare the relative intensity of the reflected light in the second preselected wavelength range detected at each sensor element in the two-dimensional matrix of sensor elements to an average relative intensity of the reflected light in the second preselected wavelength range detected by the two-dimensional matrix of sensor elements. [0023] In some embodiments, each sensor element may comprise a first sub- element sensitive to the first preselected wavelength range and a second sub-element sensitive to the second preselected wavelength range.

[0024] In some embodiments, the first preselected wavelength range can be in the BV spectrum, and the second preselected wavelength range can be in the ROY spectrum.

[0025] In some embodiments, the reflected light in the first preselected wavelength range may comprise substantially monochromatic light with a wavelength less than 500 nm, and the reflected light in the second preselected wavelength range may comprise substantially monochromatic light with a wavelength greater than 500 nm.

[0026] In some embodiments, the image sensor can be configured to detect ambient light reflected by the material sample, and the processor can be configured to use the detected ambient light reflected by the material sample to reduce noise when computing the relative intensity.

[0027] In some embodiments, the apparatus may further comprise: a first lens element, the first lens element for focusing light reflected by the material sample onto the image sensor. In some embodiments, the apparatus may further comprise a second lens element, the second lens element for diffusing light from the at least one light source onto the material sample. The second lens element may be an end of an optical fiber

[0028] In some embodiments, the at least one light source may comprise a first Light Emitting Diode (LED) or laser source for transmitting light in the first preselected wavelength range and a second LED or laser source for transmitting light in the second preselected wavelength range. In some other embodiments, the at least one light source can be a broad spectrum light source, and the apparatus may further comprise a first filter element to filter the reflected light to isolate the first preselected wavelength range, and a second filter element to filter the reflected light to isolate the second preselected wavelength range.

[0029] In some embodiments, the image sensor can be selected from the group consisting of a CCD image sensor and a CMOS image sensor. [0030] In some embodiments, the apparatus may further comprise a portable enclosure, the portable enclosure housing the at least one light source, the image sensor and the processor.

[0031] In some embodiments, the at least one light source is configured to illuminate the material sample with light in a third preselected wavelength range, and wherein the image sensor is configured to detect the light reflected by the material sample in the third preselected wavelength range, and wherein the processor is further configured to compare intensity of the reflected light in the third preselected wavelength range to intensity of the reflected light in the first and second preselected wavelength ranges.

[0032] In some embodiments, the first preselected wavelength is in the blue-violet spectrum, the second preselected wavelength is in the red-orange-yellow spectrum, and the third preselected wavelength is in the green spectrum.

[0033] In some embodiments, the image sensor comprises a two-dimensional matrix of sensor elements.

[0034] In some embodiments, the processor is configured to identify a reflected light intensity in the second preselected wavelength range greater than 55% of the maximum intensity.

[0035] In some embodiments, the comparison performed by the processor comprises computing a reflectance slope product, wherein the reflectance slope product is a result of the square of reflected light intensity in the second wavelength multiplied by reflected light intensity in the third wavelength, and divided by the square of reflected light intensity in the first wavelength.

[0036] In some embodiments, the apparatus further comprises a focusing lens element, the focusing lens element for focusing light reflected by the material sample onto the image sensor.

[0037] In some embodiments, the apparatus further comprises at least one optical fiber coupled to the at least one light source at an input end of the at least one optical fiber, the at least one optical fiber having an opposite, output end positioned to illuminate the material sample. [0038] In another broad aspect, there is provided an apparatus for detecting gold in a material sample, the apparatus comprising: at least one light source for illuminating the material sample with light in a first preselected wavelength range less than 500 nm and with light in a second preselected wavelength range greater than 500 nm; an image sensor comprising a two-dimensional matrix of sensor elements for detecting the light in the first preselected wavelength range reflected by the material sample and the light in the second preselected wavelength range reflected by the material sample; and a processor for computing a reflectance ratio product for each sensor element in the two- dimensional matrix of sensor elements, wherein the reflectance ratio product comprises the product of a wavelength reflectance ratio and a spatial reflectance ratio, wherein the wavelength reflectance ratio comprises relative intensity of the reflected light in the second preselected wavelength range to the reflected light in the first preselected wavelength range, and wherein the spatial reflectance ratio is computed based on the ratio of relative intensity of the reflected light in the second preselected wavelength range in each individual sensor element to relative intensity of the reflected light in the second preselected wavelength range in a plurality of sensor elements.

[0039] In yet another broad aspect, there is provided an apparatus for detecting gold in a material sample, the apparatus comprising: at least one light source for illuminating the material sample with light in a first preselected wavelength range, with light in a second preselected wavelength range and with light in a third preselected wavelength range, wherein the first preselected wavelength is in the blue-violet spectrum, wherein the second preselected wavelength is in the red-orange-yellow spectrum, and wherein the third preselected wavelength is in the green spectrum; an image sensor comprising a two-dimensional matrix of sensor elements for detecting the light in the first, second and third preselected wavelength ranges reflected by the material sample; and a processor for computing a reflectance slope product, wherein the reflectance slope product is a result of the square of reflected light intensity in the second wavelength multiplied by reflected light intensity in the third wavelength, and divided by the square of reflected light intensity in the first wavelength.

[0040] In a further broad aspect, there is provided a non-transitory computer- readable storage medium with an executable program stored thereon. The executable program may be configured to instruct a processor to perform a method for detecting gold in a material sample, the method comprising: illuminating the material sample with light in a first preselected wavelength range and with light in a second preselected wavelength range; detecting the light in the first preselected wavelength range reflected by the material sample and the light in the second preselected wavelength range reflected by the material sample; and comparing a relative intensity of the reflected light in the second preselected wavelength range to the reflected light in the first preselected wavelength range.

DRAWINGS

[0041] For a better understanding of the embodiments described herein and to show more clearly how they may be carried into effect, reference will now be made, by way of example only, to the accompanying drawings which show at least one example embodiment, and in which:

FIG. 1 is an example plot of optical reflectance as a function of wavelength for gold (Au), silver (Ag) and aluminum (Al);

FIG. 2A is an example plot of Reflectance Ratio Product against gold particle diameter for gold hosted in pyrite;

FIG. 2B is an example plot of Reflectance Ratio Product against gold particle diameter for gold hosted in chalcopyrite;

FIG. 2C is an example plot of Reflectance Ratio Product against gold particle diameter for gold hosted in arsenopyrite;

FIG. 2D is an example plot of Reflectance Ratio Product against gold particle diameter for gold hosted in pyrrhotite;

FIG. 2E is an example plot of Reflectance Ratio Product against gold particle diameter for gold hosted in sphalerite;

FIG. 2F is an example plot of Reflectance Ratio Product against gold particle diameter for gold hosted in galena;

FIG. 2G is an example plot of Reflectance Ratio Product against gold particle diameter for gold hosted in quartz;

FIG. 3A is an example flow diagram of a gold detection process in accordance with some embodiments;

FIG. 3B is an example flow diagram of a signal processing algorithm in accordance with some embodiments; FIG. 4 is a schematic diagram of an example apparatus in accordance with some embodiments;

FIG. 5 is a schematic diagram of another example apparatus in accordance with some embodiments;

FIG. 6 is a diagram of an example portable enclosure for the apparatus according to some embodiments;

FIG. 7 is a plot of example reflectance data for a sample of gold ore;

FIG. 8A is an example flow diagram of a detection process;

FIG. 8B is an example flow diagram of a signal processing algorithm that may be carried out by a processor in some embodiments;

FIG. 9 illustrates an example apparatus for detecting gold in accordance with some embodiments;

FIG. 10 is a plan view illustrating an example placement of optical fibers around a sample area in the example apparatus of FIG. 9; and

FIG. 11 illustrates another example apparatus for detecting gold in accordance with some embodiments.

[0042] The skilled person in the art will understand that the drawings, described below, are for illustration purposes only. It will be appreciated that for simplicity and clarity of illustration, elements shown in the figures have not necessarily been drawn to scale. For example, the dimensions of some of the elements may be exaggerated relative to other elements for clarity. Further, where considered appropriate, reference numerals may be repeated among the figures to indicate corresponding or analogous elements.

DESCRIPTION OF VARIOUS EMBODIMENTS

[0043] A basic physical property of gold is its high optical reflectance in the red, orange and yellow part of the visible spectrum (i.e., where the wavelength of light is greater than 500 nm) and its relatively lower optical reflectance in the blue and violet part of the visible spectrum (i.e., where the wavelength of light is lower than 500 nm).

[0044] The optical reflectance of gold can be lower by a factor of about three in blue and violet (BV) light, as compared to its reflectance in red, orange and yellow (ROY) light. This relatively unique optical property is what gives gold its characteristic brilliant colour, and it is a consequence of relativistic effects in the outer electron orbitals of the gold atom. Such relativistic effects account for the rapid change in the reflectance of gold and silver as a function of the wavelength of incident light. These relativistic effects are described in further detail by Pyykko, P. and Desclaux, J. P., "Relativity and the Periodic System of Elements", Journal of the American Chemical Society, Accounts of Chemical Research, 1979, 12, 276; Pyykko, P., "Relativistic Effects in Structural Chemistry", Chemical Review, 1988, 88, 563-594; and Schwerdtfeger, P., "Relativistic Effects in Properties of Gold", Heteroatom Chemistry, 2002, Vol. 3, No. 6, 578-584, the contents of which are incorporated herein by reference.

[0045] Referring now to FIG. 1 , there is shown a graph of optical reflectance as a function of wavelength for three metals: gold (Au), silver (Ag) and aluminum (Al). A steep change in reflectance for gold at approximately 500 nm can be observed. A similar change in reflectance can also be observed for silver at about 325 nm, in the ultraviolet spectrum. However, the optical reflectance for aluminum remains relatively constant at all the wavelengths depicted. The principles of relativistic quantum chemistry that give rise to the reflectance characteristics displayed in FIG. 1 are known in the art.

[0046] These reflectance properties can be similar in an alloy, such as a gold and silver alloy, which often occurs in nature. There is no chemical bond between the gold and silver in such an alloy. In the absence of chemical bonds, the behavior of the respective outer electron orbitals of the atoms in the alloy is unchanged. Accordingly, the optical reflectance properties of the respective atoms are also unchanged.

[0047] However, in practice, the reflectance properties of a gold and silver alloy will have a reflectance transition point at a wavelength that is intermediate between the 325 nm rise for silver and the 500 nm rise for gold. The precise transition point will depend on the relative quantities of the respective constituent metals. A higher concentration of gold will result in a transition closer to the 500 nm transition point of gold. Conversely, a lower concentration of gold will result in a transition closer to the 325 nm transition point of silver. Typically gold and silver alloys predominately consist of gold. Accordingly, the rise is likely to be at a higher wavelength (e.g., in the 450 nm to 500 nm range).

[0048] Conventionally, in the geological community, it is believed that gold occurs at such low concentrations in mineral samples in nature, that there is no method to detect gold in mineral samples in the field using a hand held device. [0049] However, by taking advantage of the above-described optical reflectance properties, even small gold particles can be detected using the techniques described herein. As long as a gold particle, however small, is close enough to the face of a mineral or material sample (e.g., a cleaved rock face) to reflect light, then its characteristic reflectance properties should be detectable.

[0050] In particular, by comparing the optical reflectance properties for a mineral or material sample when exposed to wavelengths in the BV range to the properties when exposed to wavelengths in the ROY range, gold and gold-silver alloys can be readily distinguished from other metals and minerals. It will be understood that a mineral sample need not be comprised entirely of a single mineral and may comprise other materials.

[0051] For example, if a ROY light source is chosen with a wavelength of 700 nm (red) and directed at a gold particle, the reflectance will be approximately 84%. If a BV light source is chosen with a wavelength of 400 nm (violet) and directed at the gold particle, the reflectance will be approximately 25%. Accordingly, the ratio of the gold particle's reflectance at the different wavelengths is 84% divided by 25%, or approximately 3.4

[0052] In general, to detect gold, the ROY light source should be chosen to have a wavelength of greater than 500 nm and the BV light source should be chosen to have a wavelength of less than about 500 nm, and preferably in the range of 325 nm to 500 nm. Gold-silver alloys may also be detected using similar techniques. Native silver may be more difficult to detect, as it may be tarnished with an oxide and/or sulphide coating and be relatively unreflective. However, it is possible to detect freshly cleaved native metallic silver that has not yet tarnished by employing a light source with a wavelength less than 325 nm, which corresponds to the BV wavelength, and a light source with a wavelength greater than 325 nm that corresponds to the ROY wavelength. Alternatively, it is possible to detect freshly cleaved native metallic silver that has not yet tarnished by employing a light source with a wavelength less than 325 nm, that corresponds to the BV wavelength, and a light source with a wavelength greater than 325 nm that corresponds to the Green wavelength, and a light source with a wavelength in the ROY part of the visible spectrum. [0053] Table 1 provides reflectance values measured using optical techniques for a variety of materials - including gold, pyrite, chalcopyrite, arsenopyrite, pyrrhotite, sphalerite, galena and quartz - in a range of wavelengths between 400-700 nm.

[0054] In some cases, multiple forms of a material may be present in a sample. For example, there may be equal parts of arsenopyrite R1 and arsenopyrite R2. Accordingly, to obtain the reflectance for the mixture, the reflectance value for each constituent component at the particular wavelength may be averaged. In most cases, gold particles will occur embedded in a mineral or other material sample among one or more minerals and other materials. Accordingly, the overall or average reflectance of the entire mineral or material sample may not exhibit the properties particular to gold. Reliably and efficiently detecting small particles of gold in such a sample may therefore require obtaining an array of localized reflectance measurements. [0055] In some embodiments, the array of measurements can be obtained by using an image sensor, such as a complementary metal oxide semiconductor (CMOS) active pixel sensor device, or a charge-coupled device (CCD) image sensor. Such image sensors typically comprise a two-dimensional grid or array of photosensitive elements, which are used to generate a two-dimensional raster image, which comprises a corresponding grid. The raster image comprises a plurality of pixels corresponding to each point on the grid.

[0056] By directing substantially monochromatic light at a mineral sample (or by filtering images for particular desired wavelengths of light), and using a suitable image sensor to image the mineral sample while exposed to the monochromatic light, a relatively large array of reflectance sample points can be obtained in an efficient manner. In particular, an image sensor may comprise a two-dimensional matrix of photosites or sensor elements, yielding a large number of individual sample points. The individual photosites or sample points will correspond to pixels in an image generated by the image sensor. Photosites in the resulting image that appear "bright" can be interpreted as having a high average ROY reflectance, whereas photosites that appear dark may have a low average ROY reflectance. A photosite is a group of photo sensors generally comprising at least one each of red, green and blue photo sensors. The most common type of photosite is based on the Bayer filter, and has one photo sensor configured to detect red light, two photo sensors configured to detect green light and one photo sensor configured to detect blue light. Color filters may be used to configure each photo sensor to detect the desired range of light. However, other configurations of photo sensors and/or color filters may also be used. For example, a Foveon™ image sensor has a three dimensional integration of photo sensitive elements: one for detecting red light, one for detecting green light and one for detecting blue light.

[0057] Accordingly, in some embodiments, the sensor elements may comprise two or more sub-elements. For example, a color-sensitive sensor element may comprise two or more sub-elements that are sensitive to specific wavelength ranges of light. Alternatively, the sensor element may comprise two or more sub-elements that are provided with filters that correspond to specific wavelength ranges of light (e.g., red, blue, etc.).

[0058] Accordingly, the reflectance sample points obtained in this way can be compared against known reflectance values, such as shown in Table 1 , to generally classify the materials that appear in the image. For example, a photosite that is dark when exposed to light with a wavelength of 700 nm can be ruled out as corresponding to gold.

[0059] By taking multiple exposures of a mineral sample when exposed to different wavelengths of light, desired reflectance characteristics as a function of wavelength can be identified.

[0060] For example, in some embodiments two or more light sources are used. Laser or LED light sources are suitable because they can produce substantially monochromatic light at a desired wavelength. A first laser or LED, emitting violet light at approximately 400 nm can be used to illuminate an area (e.g., one square centimeter) of the mineral sample and a first image captured of the light reflected by the material sample. The process may be repeated with a second laser or LED emitting red light at approximately 650 nm. Using a color sensor element with filters over the elements that transmit only red or blue-violet light, the red and violet light can be detected simultaneously by the red-filtered element and blue-violet-filtered element respectively, in a color sensor element.

[0061] In some embodiments, the mineral sample is shielded from external light. However, if shielding is not absolute, or to improve accuracy, a third image may also be captured without illumination by either the first or second laser or LED light source, to serve as a reference image.

[0062] The captured images can be processed and compared on a photosite-by- photosite basis to analyze the reflectance as a function of wavelength and position within the photosite array. Individual photosites in each of the captured images can be aligned to correspond to the same point of the mineral sample. Accordingly, differences in the brightness of the aligned photosites can be compared and used to identify the presence of gold or other materials.

[0063] To detect particular materials, such as gold, a weighted intensity value may be calculated for each photosite. In some embodiments, the weighted intensity value may comprise a Wavelength Reflectance Ratio (WRR). The Wavelength Reflectance Ratio may be a simple ratio of the photosite intensity in the second image (e.g., under ROY light) versus the photosite intensity in the first image (e.g., under BV light). Equation (1 ) can be used when the ROY and BV light intensities that are emitted from the optical fiber 420 in FIG. 4 are arranged to be equal so that the ROY and BV incident light intensities cancel when the ROY and BV reflectances are divided to calculate the Wavelength Reflectance Ratio. If the ROY and BV incident light intensities are not equal, then a factor equal to the ratio of the ROY and BV incident light intensities can be included in equation (1 ). In practice, the ROY and BV incident intensities can be set to be equal because this makes the calculation of WRR simpler. However, as noted, this does not need to be the case if a factor is introduced to account for the difference in ROY and BV incident intensities.

[0064] For a photosite occupied entirely by gold, the Wavelength Reflectance Ratio will have a value of approximately 3 if the intensity of the incident ROY and BV light is the same, and depending upon the specific wavelengths of ROY and BV light used. Due to its high reflectance for ROY light, a photosite with a significant fraction of its area occupied by gold is called a "bright photosite".

[0065] A Spatial Reflectance Ratio (SRR) can also be calculated for the bright photosites. The Spatial Reflectance Ratio can be the ROY reflectance for a bright photosite, which may be a candidate to contain gold, divided by the average ROY reflectance for all non-bright photosites in the image.

[0066] The Spatial Reflectance Ratio for a particular photosite provides a measure of how bright that particular photosite is, or how much it "glitters" as compared to the host material.

[0067] For a photosite that captures only the host mineral, the SRR should be approximately 1 , as the ROY reflectance will be the same or similar to the average of the ROY reflectance of the photosites containing host minerals.

[0068] The Wavelength Reflectance Ratio can be used to help discriminate gold from other reflective materials, such as sulphides, whose reflectance does not exhibit the large difference in reflectance between ROY and BV light.

[0069] In some embodiments, a product can be computed by multiplying the Wavelength Reflectance Ratio by the Spatial Reflectance Ratio. This product can be considered the "Reflectance Ratio Product" (RRP), which can serve as a useful indicator of whether gold is present in the area of the mineral surface viewed by a given photosite. The measurement sensitivity of the Reflectance Ratio Product can make it possible to identify gold particles in the submicron and micron size range.

[0070] For a photosite that captures only the host mineral, the WRR will be the ratio of the intensities at ROY and BV wavelengths for the host mineral, which can be more or less than 1. The SRR, as explained above, can be approximately 1. Accordingly, the RRP for the host mineral can be more or less than 1. Values of the RRP for a photosite filled with host mineral, and a photosite filled with host mineral and a gold particle, can be seen in Tables 2A to 2G herein. The percentage of the RRP for a photosite containing a gold particle above that of the RRP for a photosite containing only host mineral can also be seen.

[0071] As an example, using an image sensor with a resolution of 1920x1080 photosites (e.g., a total of 2,073,600 photosites), a single pass of the above-described technique can perform the equivalent of 2,073,600 measurements in parallel, as each photosite in the image corresponds to an individual measurement. If a single photosite corresponds to an area of 50 square microns of the mineral sample, an image comprising 2,073,600 photosites would correspond to approximately 1 square centimeter of the mineral sample. Commonly available CMOS and CCD image sensors have between ten to twenty million photo sensors, which corresponds to between two and a half and five million photosites, since there are typically four photo sensors per photosite. Detection resolution may also be expected to significantly improve with greater pixel capacity of image sensors, and with the ability to focus the image on smaller areas.

[0072] Typical gold grades encountered in the field can be of the order of 1 part per million. Thus, performing millions of measurements in parallel may enable gold to be detected, if present, with only a few imaging steps.

[0073] Gold particles may be detectable even if they do not fill an entire photosite of the resulting image. The reflectance properties of the gold may be "muted" by the surrounding material, but the overall reflectance for the photosite may be sufficient to facilitate detection. For example, given a photosite size corresponding to 50 square microns of the mineral sample, even gold particles in the submicron range may still provide a detectable increase in the Reflectance Ratio Product for that photosite.

[0074] Tables 2A to 2G describe the effect of gold particle size on a calculated Reflectance Ratio Product in a variety of materials, given a photosite area corresponding to 50 square microns of the mineral sample. Gold particles were modeled as a circle, and the measured reflectance data as a function of wavelength for gold and a variety of host minerals is tabulated, along with the computed reflectance ratios and products.

[0075] Photosites containing a gold particle can exhibit a Reflectance Ratio Product that exceeds that of photosites containing only the host material. A detection threshold may be used to eliminate spurious detections and false positives. For example, the detection threshold may be specified to require a 10% higher Reflectance Ratio Product than the known or measured Reflectance Ratio Product of the host material.

[0076] Using the criterion that the Reflectance Ratio Product for a photosite containing a gold particle must exceed the Reflectance Ratio Product for a photosite containing only the host mineral by 10%, Table 2A illustrates that a gold particle 2.3 microns in diameter can be detected in pyrite host material. For unreflective host minerals like quartz, shown in Table 2G, particles as small as 0.5 microns in diameter can be detected.

[0077] Accordingly, photosite-by-photosite comparison of the Reflectance Ratio Product may permit the identification of gold particles on the scale of microns.

[0078] In many cases, gold is likely to be present in only trace quantities. As several million photosites can be measured at once using an image sensor, and as gold is likely to be present in only a few photosites, the photosites that do not initially exhibit one or more reflectance properties of gold can be used to provide a large sample of the reflectance properties of the host material. The reflectance properties of the host material can be estimated by computing an average reflectance of the non-bright photosites that do not appear to contain gold.

[0079] Accordingly, in a first stage of analysis, the reflected ROY light intensity of each photosite can be measured and compared to the average ROY light intensity reflected by all photosites (e.g., Spatial Reflectance Ratio). The Spatial Reflectance Ratio can be considered as a measurement that detects bright photosites that are likely to contain gold.

[0080] In a further stage, for photosites with higher than average reflectance of ROY light (e.g., bright photosites), the Wavelength Reflectance Ratio for these photosites can be computed (e.g., for ROY light to BV light). Depending on the specific ROY and BV wavelengths used, the ratio will be approximately 3 if the photosite is entirely occupied by gold.

[0081] Referring now to FIG. 3A, there is shown an example flow diagram of a detection process in accordance with some embodiments.

[0082] A mineral sample may be obtained and cleaved or cut with a diamond saw to reveal a generally smooth surface. Cutting with a diamond saw is the conventional method for splitting diamond drill core samples, because half of the sample can be sent to a conventional assay office and half can be retained at the mine or exploration site. Cutting a mineral sample with a diamond saw provides a surface that is flat enough for measurement as described herein. The diamond saw also cuts through sulphide minerals such as pyrite, chalcopyrite, arsenopyrite, pyrrhotite and sphalerite that can encapsulate gold particles, thereby making the gold particles visible and detectable. Conventionally, in geology, it is not believed that gold particles encapsulated in sulphide minerals can be observed optically, however the cutting process can reveal gold.

[0083] Detection process 300 begins at 305 by illuminating a material sample with a first preselected wavelength range of light. The light can be directed such that it reflects from a desired surface of the mineral sample, and the reflected light can be detected at 310, for example by an image sensor to produce a first image.

[0084] Similarly, at 315, the material sample can be illuminated with a second preselected wavelength range of light. The light can again be directed such that it reflects from a desired surface of the mineral sample, and the reflected light can be detected at 320, again for example by an image sensor to produce a second image.

[0085] The first preselected wavelength range of light may be in the BV spectrum and the second preselected wavelength range of light may be in the ROY spectrum, although it will be appreciated that these may be reversed.

[0086] In some embodiments, 305 and 315 may be performed substantially simultaneously. For example, rather than using separate light sources, each producing a distinct wavelength range of light (e.g., monochromatic lasers or LEDs), a broad spectrum light source can be used to produce a wide wavelength range of light, comprising both the first and second wavelength ranges. Accordingly, the material sample can be illuminated with both ranges simultaneously. In order to detect the first and second wavelength ranges separately, first and second filter elements may be used to isolate the first and second wavelength ranges, respectively.

[0087] For example, 305 and 315 may be performed substantially simultaneously using monochromatic light produced by lasers or LEDs that are projected together, and detected together on a color photosite containing elements with color filters that permit one element to detect only ROY light and the other to detect only BV light. Such an arrangement can be seen in FIG. 4 where the ROY and BV light can be simultaneously coupled into an optical fiber and then emitted together from the core of the optical fiber, and in FIG. 5 where the ROY and BV light are detected simultaneously by a color photosite with a red filter over one element and a blue filter over another element. The filter materials can be selected to transmit the ROY and BV light through each of their respective filters.

[0088] Alternatively, 305 and 315 may be performed sequentially, in which case the color filters may not be required.

[0089] Optionally, at 325, the at least one light source may be disabled and ambient light may be detected to generate a reference image. At 330, the images or data may be processed to eliminate the effect of ambient light.

[0090] Referring now to FIG. 3B, there is shown an example flow diagram of a signal processing algorithm that may be carried out by a processor in some embodiments to analyze intensity and relative reflectance data. [0091] Analysis process 340 begins at 355, by retrieving image data from the image sensors. The image data can comprise a ROY image and a BV image, where each image is generated by exposing a sample to substantially monochromatic light at a suitable wavelength (e.g., 635 nm for ROY and 405 nm for BV).

[0092] At 360, the processor can discard or ignore photosites with a reflected light intensity that is below a predetermined exposure threshold value. The predetermined exposure threshold value can be used to discard image data that is underexposed or unexposed, for example where an area of the sample was not illuminated by a light source.

[0093] The average ROY light intensity for the ROY image can be computed at 365 by averaging the reflected light intensity of each illuminated photosite (e.g., each photosite that was not discarded at 360). The average ROY light intensity can be further refined by eliminating photosites from the average whose ROY light intensity exceeds the first calculated average by a predetermined amount (e.g. 10%). This provides an average ROY light intensity for photosites that are unlikely to contain gold.

[0094] A normalized ROY light intensity called the Spatial Reflectance Ratio (SRR) for each photosite in the ROY image can be computed at 370 by dividing each photosite's ROY light intensity by the average ROY light intensity, as computed at 365.

[0095] Photosites with a Spatial Reflectance Ratio that exceeds a minimum threshold can be selected at 375. For example, the processor may identify those photosites whose Spatial Reflectance Ratio exceeds the average Spatial Reflectance Ratio by a predetermined amount or percentage (e.g., 10%). The identified photosites may be considered as "bright photosites".

[0096] For each of the photosites the processor can divide the ROY light intensity for the photosite by the BV light intensity for the photosite at 380, to compute the Wavelength Reflectance Ratio (WRR) for the photosite.

[0097] At 385, the value of the Reflectance Ratio Product (RPP) for each photosite may be computed by multiplying the Wavelength Reflectance Ratio (WRR) by the Spatial Reflectance Ratio (SRR) as specified by equation (3).

[0098] A Reflectance Ratio Product for a bright photosite greater by a certain amount (e.g. 10%) than the average Reflectance Ratio Product for non-bright photosites, may indicate the presence of gold in a given bright photosite. [0099] Based on the Reflectance Ratio Product value, the processor may determine whether gold is present in a given photosite at 390. If gold is present, the Wavelength Reflectance Ratio may further be used to estimate the diameter of the gold particle or particles in each bright photosite at 395. This can be done using equation (4):

[00100] Where x is the fraction of the photosite covered by gold. R _H BV can be the reflectance of the host mineral for the BV wavelength measured from the average BV reflected light intensity and the incident light intensity calculated from the laser or LED light power, the area illuminated and the area of the photosite. R _H BV can also be determined from tabulated data if the user is able to supply the type of host mineral.

[00101] RHROY can be the reflectance of the host mineral for the ROY wavelength measured from the average ROY reflected light intensity and the incident light intensity calculated from the laser or LED light power, the area illuminated and the area of the photosite. RHROY can also be determined from tabulated data if the user is able to supply the type of host mineral.

[00102] RA _U BV can be the reflectance for gold at the BV wavelength which is a known quantity. It can be adjusted to account for impurities such as silver or copper if the concentrations of these impurities are known for the sample from other techniques.

[00103] RA _U ROY can be the reflectance for gold at the ROY wavelength which is a known quantity. It can be adjusted to account for impurities such as silver or copper if the concentrations of these impurities are known for the sample from other techniques.

[00104] Using the fraction x of each bright photosite that is covered by gold, the fraction of the total analyzed mineral surface covered by gold (Gold Fraction) may be calculated. Assuming that the surface area density and volume density are the same, and correcting for the difference in density between gold, or the gold alloy that is present in a given sample, and the host mineral, the gold grade can be calculated from equation (5):

[00105] In some embodiments, if gold is suspected in a predetermined sample area, or if the result for a subarea of the sample is inconclusive, the desired area may be reimaged with optical magnification applied (e.g., using a zoom lens to focus on the desired area) and the process repeated, to obtain even greater resolution and detection sensitivity. With optical magnification applied, each photosite may correspond to an area of the mineral sample that is even smaller than 50 square microns, with the concomitant increase in detection sensitivity.

[00106] Using process 340, gold particles on the order of 1 micron in diameter can be detected, and at grades as low as 50 to 100 ppb, which can be as much as three orders of magnitude more sensitive than current analyzers.

[00107] This is because the gold detector can detect a circular gold particle whose diameter is 1 micron on a 1 square centimeter area of a mineral whose area is 10 ^Λ 8 um ^A 2, and the density of naturally occuring gold alloy can be estimated to be 18 g/cm ^A 3 (with impurities reducing its density from the 19.3 g/cm ^A 3 density of pure gold) and the density of a typical host mineral is 2.7 g/cm ^A 3. Therefore the minimum detectable gold grade can be estimated as:

[00108] Typical gold grade detection limits for portable XRF detectors are in the range of 50 parts per million (ppm) in the field, and approximately 5 ppm under carefully calibrated lab conditions.

[00109] In addition, more than two light wavelengths may be used, as noted above. The additional wavelengths can be used to further enhance the ability to discriminate for gold or other materials, if required.

[00110] In an alternative embodiment, three or more images may be generated. First and second images would correspond to ROY and BV illumination as described herein. However, a third image could be generated in which neither ROY nor BV illumination is used. Accordingly, image data from the third, unexposed image in particular can be used to enhance detection quality and sensitivity, for example, by eliminating "stuck" photosites that appear in all three images, to reduce noise, or to compensate for the presence of broad spectrum visible light.

[0011 1] Referring now to FIG. 4, there is shown an example apparatus for detecting gold in accordance with some embodiments. Apparatus 400 comprises one or more light sources, for example, a first light source 405 and a second light source 410, which can be substantially monochromatic. In some embodiments, first light source 405 may be a ROY diode laser or LED and second light source may be a BV diode laser or LED. In other embodiments, there may be only a first light source 405, which produces broad spectrum light.

[00112] Examples of suitable light sources that are relatively inexpensive and readily available commercially include the indium gallium nitride (InGaN) diode lasers, such as those used in Blu-ray™ disc players, which produce light with a wavelength of 405 nm. Diode lasers such as those used in commercially-available red laser pointers can produce light with a wavelength of between 635 and 670 nm. In addition, diode lasers used in DVD™ players produce light with a wavelength of 657 nm. LEDs can also be used although the bandwidth of the light they produce is 20 nm to 30 nm, as compared to 2 nm to 3 nm for diode lasers.

[00113] The first and second light sources may be used substantially simultaneously or sequentially to illuminate a mineral sample 450 with two wavelengths of light. Light from each source may be directed to a desired area by, for example, an optical fiber 420. The light emitted from the optical fiber can be passed through a focusing lens element 425, which can be concave to spread the emitted light X over a desired area of the mineral sample 450. In some embodiments, the focusing lens element 425 may simply comprise the end of an optical fiber, as light emitted from an optical fiber may spread out from the end of the optical fiber enough to project light over a sample area of interest. Reflected light Y can be captured by another focusing lens element 430, which can be convex to focus the reflected light onto an image sensor 435.

[00114] Once an image is formed by image sensor 435, it can be transmitted to a processor 460 for further processing and/or storage in storage device 465. Processor 460 may be a general purpose processor, a field programmable gate array (FPGA) or other suitable processor, which can be programmed to perform the desired image processing.

[00115] Each of the components of apparatus 400 can be chosen to be small, light and to require low power. Accordingly, apparatus 400 can be constructed in a rugged, hand held enclosure, to act as a battery powered instrument, useful in the field for mineral exploration, and for grade control during gold mining. A hand held apparatus would require minimal training to use, and would be considerably safer to use than an X-

Ray Fluorescence (XRF) detector. In particular, apparatus 400 employs only visible light and not high energy ionizing radiation such as X-rays. Apparatus 400 could be used safely with minimal precautions to avoid directing laser light into a user's eyes. If LEDs are used as light sources, then minimal or even no precautions would be required. In addition, safety features can be incorporated into the apparatus 400 to minimize the risk of laser injury, for example a proximity sensor to ensure that lasers are only enabled if a mineral sample is within a predetermined range of the apparatus. This can ensure that the laser light is not projected at a longer distance that could damage a bystander's eyes.

[00116] In some alternative embodiments, one or more broad spectrum light sources could be used instead of monochromatic light sources. The broad spectrum light can be filtered using two or more selective wavelength filter elements as desired to generate images that correspond to ROY and BV bandwidths of light.

[00117] Referring now to FIG. 5, there is shown an example schematic diagram of a detection apparatus in accordance with both the previously described monochromatic diode laser and LED light sources, and an alternative embodiment using broad spectrum light sources. As previously described, when using ROY and BV monochromatic diode laser and LED light sources, the ROY and BV light can be detected by the color photosite simultaneously. This can be accomplished using the red and blue filters over their respective sensor elements. The red filter transmits only the ROY light, and the blue filter transmits only the BV light. Therefore the ROY and BV light can be detected simultaneously by the color photosite without interfering with each other.

[001 8] Alternatively, if a single white light source is projected and reflected off a mineral sample, and detected by a color sensor array, then the red and blue light can be filtered by the color sensor elements, and may be used to generate the desired images for carrying out process 300.

[00119] Accordingly, apparatus 500 can comprise a broad spectrum light source 507, which projects light X' onto a mineral sample 550. Reflected light Y' can be detected by image sensor 536, which in this case is a color array. The image sensor may have color filter elements 537, which generate the desired reflectance image.

[00120] The color filter approach can have the advantage that it is relatively simple and inexpensive to produce. For example, the color filter for a color photo sensor may be patterned lithographically over the sensor elements during the fabrication process. However, accuracy may be reduced, since the red and blue filters in an element may admit light in approximately a 50 nm to 100 nm bandwidth range. In contrast, the bandwidth of a diode laser may be only a few nanometers, while the bandwidth of an LED is generally between 20 nm and 30 nm. The wider bandwidth of the filtered light may result in diminished detection capability for the sharp change in the reflectance characteristic of gold at 500 nm. Accordingly, this may result in a larger minimum gold particle size that can be identified and a higher minimum gold grade that can be measured.

[00121] As white light sources are generally considerably more diffuse than laser or LED sources, no lenses are required in apparatus 500. However, if desired, both convex and concave lenses may also be used in a similar manner as in apparatus 400.

[00122] Referring now to FIG. 6, there is shown a diagram of an example portable enclosure for a detection apparatus, such as apparatus 400, 500, 900 or 1 100. Enclosure 600 can be provided in a "pistol-grip" or other suitable enclosure.

[00123] Enclosure 600 can comprise a handle 685, a trigger actuator 690 for activating various functions of the apparatus (e.g., capturing images for detection), a control panel 665, display 670 for interacting with software in the apparatus to configure various settings, a speaker 675 to provide an audible signal for the detection of gold, an on/off switch 680, a flexible seal 660 for covering a mineral sample under examination, to block out unwanted light, and an antenna 695, which may be used for wireless communication of, for example, detection data.

[00124] Display 670 may display information regarding gold grade, gold particle size and other information. Likewise, control panel 665 may provide settings for the operation of the apparatus, such as customized analysis programs for particular host minerals. In some embodiments, display 670 and control panel 665 may be combined in a touch screen device.

[00125] In some embodiments, computation of the average ROY reflected intensity can be omitted in the measurement of the spatial reflectance ratio by adjusting red light intensity such that a maximum range of red intensities is detected. For example, the intensity of a red light source may be increased from a minimum until saturation of a small proportion of red pixels (e.g., between 1 to 100 out of one to ten million pixels) is achieved. At the same time, green and blue light intensities also may be increased proportionately with the red light intensity. By saturating a small number of red pixels, maximum use of the dynamic range of the CMOS or CCD sensor can be made. By saturating a few red pixels, the range of red intensities is generally maximized, and therefore an absolute measure of red intensity can be used to measure the spatial reflectance ratio rather than a relative measure that is relative to an average red intensity. Thus it is not necessary to calculate the average red intensity, and the red intensity can replace the spatial reflectance ratio in calculations of the reflectance ratio product.

[00126] In such embodiments, green light intensity may also be used to assist in identifying photosites containing gold, where a photosite contains at least one each of red, green and blue photo sensors. The use of green light intensity facilitates the detection of photosites with an intensity spectrum that rises steeply with wavelength through the visible light range, which is characteristic of gold.

[00127] Generally, a sensor element detects a single colour such as red, green or blue by using a filter material deposited over the photodiode of the pixel. A sensor element, in a Bayer filter CMOS or CCD sensor, consists of a group of four pixels: one red, two green and one blue. In a Foveon™ CMOS sensor a sensor element consists of one red, one green and one blue pixel, where the red, green and blue pixels are stacked vertically using two layers of n-type silicon and two layers of p-type silicon, that alternate n-type and p-type. The red, green and blue intensities measured by the pixels for a given photosite record the colour of the incident light. When a photosite is used to observe a gold particle, the colour the photosite records is a golden colour, which has a high red intensity, a medium green intensity, and a low blue intensity. The approximate ratios of the red to green to blue intensities for gold are red:green:blue = 3:2:1.

[00128] For example, a useful condition to select photosites that could potentially contain an image of gold is to identify photosites in which the red light intensity is at least 45% greater than the green light intensity, and the green light intensity is at least 50% greater than the blue light intensity. This is a less stringent requirement than the ratio of red.green.blue = 3:2.1 , but serves as a useful filter to select photosites for further testing. Photosites (or sensor elements) that indicate a steep rise in intensity with wavelength, and which have a high reflectance in red light (e.g., detected red light intensity is at least 55-65% of maximum intensity) may be strong candidates to contain gold. Furthermore, since gold strongly absorbs blue light, the intensity of the reflected blue light should be low if the photosite is a strong candidate to contain gold (e.g. , detected blue light intensity is no more than 25-30% of maximum intensity).

[00129] In addition, photosites that are candidates to contain gold may be further refined by computing equation (7) to obtain the Reflectance Slope Product (RSP):

where / _r , l _g and l _b are the detected intensities of red light, green light and blue light, respectively.

[00130] Equation (7) accordingly is the product of red light intensity, multiplied by the steepness of the spectrum intensity transition from red to blue and further multiplied by the steepness of the spectrum intensity transition from green to blue.

[00131] When the RSP is at least in the range 1.75 to 2.00 times the maximum intensity level (in red, green or blue), then the photosite can be selected as a strong candidate to contain gold. For example, in an 8-bit image sensor, where the maximum intensity level can be expressed as the integer number 255, the RSP should be at least 450 for a photosite to be considered a strong candidate to contain gold. In another example with a 2-bit sensor, where the maximum intensity level corresponds to integer number 4095, the RSP should be at least 7200.

[00132] Equation (7) combines the effects of relative reflectance at different wavelengths, and the spatially based relative brightness of photosites, which permits the identification of a material such as gold. Similarly, the condition wherein red light intensity is at least 55-65% of maximum is a spatially based condition, since the incident intensity of red light has been adjusted so that at least some red pixels are saturated thereby defining a scale for red intensity. - [00133] It is the combined effects of reflectance relative to wavelength and position that facilitate identification of gold, and enable distinguishing gold from sulphide minerals that are also highly reflective and have variation in reflectance with wavelength, albeit with different characteristics than gold. Gold particles are often found embedded in sulphide minerals. [00134] In some embodiments, the photosites that are determined to contain an image of gold, may contain a complete or a partial image. In some cases, these "fractional" photosites may contain, for example, only 50% gold coverage. All of the gold bearing photosites generally satisfy the condition wherein red light intensity is at least 45% greater than green light intensity, and green light intensity is at least 50% greater than blue light intensity. Likewise, these photosites may exhibit red light intensity greater than 55-65% of maximum intensity, and they satisfy the condition in equation (7).

[00135] However, the value of the RSP for each photosite may be used to estimate its fractional gold coverage. For example, an RSP value of 1.75 times the maximum intensity, may correlate to 50% gold coverage, and an RSP of 2.75 times the maximum intensity may correlate to 100% coverage. Intermediate coverage values can be determined by plotting fractional photosite coverage as determined by optical methods (e.g., scanning electron microscope imaging) and comparing with RSP values for corresponding photosites to determine an interpolation function (which may be linear, for example). RSP values in excess of 2.75 times maximum intensity generally correspond to 100% coverage. These RSP values may be higher than 2.75 times maximum intensity because a gold particle may be situated so that it reflects more light through the lens of the gold detector, and/or its surface has less material on it than other gold particles with lower RSP values, which would prevent light from being reflected.

[00136] In some embodiments, an estimated gold grade (G can be computed using equation (8):

where N _Au is the number of photosites that contain gold, NfAu is the fractional amount of photosites that contain gold, N is the total number of photosites, p _Au is the density of gold and p _HO st is the density of the host mineral. For fractional photosites, the fractional amount of gold in the photosite is summed. Conversely, for full gold photosites, the number 1 is summed.

[00137] In some cases, the estimated gold grade may be computed for a plurality of measurements (e.g., 10-30 measurements) to refine the estimate. Each measurement may constitute imaging a distinct site (e.g., 2.5mm x 4mm) on the surface of a mineral sample using all of the photosites available to the CMOS or CCD sensor.

[00138] Based on this approach, an estimate of the lower measurability limit for an example gold detector can be provided by calculating the minimum detectable gold grade when only a single photosite is detected in a single measurement out of 25 total measurements, and where that single photosite is 50% covered with gold. In this scenario, the minimum detectable gold grade G' would be only 30 ppb (0.03 g/tonne), based on a total number of photosites N = 4494540, a specific gravity for impure gold = 18, and a specific gravity for the host rock = 2.7. For comparison, the average gold grade found in the earth's crust is approximately 5 ppb (0.005 g/tonne).

[00139] If it is known that the gold in the mineral sample is not pure gold (e.g., the gold is in the form of electrum - a gold and silver alloy), then the specific gravity of the electrum alloy may be used instead of that for gold when computing estimated gold grade.

[00140] In some embodiments, the techniques described herein can be adapted to detect electrum and pure silver. In such cases, a shortest wavelength light source should be selected such that the wavelength range is low enough to detect the low reflectance region of the reflectance curve for silver (e.g., 300 nm), and the medium wavelength light source should be selected to fall within the intermediate part of reflectance curve for silver (e.g., 350 nm). A longest wavelength light source can also be used (e.g., 660 nm). The low and medium wavelength LEDs used to detect silver fall within the ultraviolet range.

[00141] Referring now to FIG. 7, there is illustrated a plot of example reflectance data for a sample of gold ore. The y-axis of plot 700 indicates the RSP for each data point, whereas the x-axis indicates the red light intensity. As shown in plot 700, light intensities are calculated using absolute values between 0-255, corresponding to an 8- bit image sensor.

[00142] In plot 700, gold photosites are indicated by '*' markers, and pyrite photosites are indicated by '-' markers. It can be observed that gold photosites have an RSP that exceeds 500 and a red light intensity greater than 145. Conversely, pyrite photosites generally have an RSP lower than 500 and a red light intensity less than 45.

[00143] The estimated gold grade was computed to be 20.79 g/tonne using equation (8), based on a specific gravity of impure gold equal to 18, and a specific gravity of the host mineral equal to 2.7. A conventional assay of the sample produced a gold grade of 20.95 g/tonne. Accordingly, the estimated gold grade was found to be generally accurate. [00144] Referring now to FIG. 8A, there is shown an example flow diagram of a detection process in accordance with some embodiments.

[00145] A mineral sample may be obtained and cleaved or cut with a diamond saw to reveal a generally smooth surface. Cutting with a diamond saw is the conventional method for splitting diamond drill core samples, because half of the sample can be sent to a conventional assay office and half can be retained at the mine or exploration site. Cutting a mineral sample with a diamond saw provides a surface that is flat enough for measurement as described herein. The diamond saw also cuts through sulphide minerals such as pyrite, chalcopyrite, arsenopyrite, pyrrhotite and sphalerite that can encapsulate gold particles, thereby making the gold particles visible and detectable.

[00146] Detection process 800 begins at 805 by illuminating a material sample with first, second and third preselected wavelength ranges of light. The light can be directed such that it reflects from a desired surface of the mineral sample. The first, second and third preselected wavelength ranges may be blue-violet (blue), red-orange-yellow (red) and green, respectively.

[00147] The reflected light in each wavelength range can be detected at 810, for example by a color image sensor with sensor elements corresponding to the respective wavelength ranges, to produce a first image.

[00148] In some embodiments, 805 may be performed using generally monochromatic light produced by lasers or LEDs that are projected at the sample surface, and detected together on a color photosite containing elements sensitive to different ranges of light. In some embodiments, a broad spectrum (e.g., white) light source may be used.

[00149] In at least some embodiments, a Foveon™ CMOS sensor is particularly well suited as a sensor for gold detection because its photosites can detect red, green and blue light that are incident on the same surface of the silicon substrate. In a Foveon™ CMOS sensor the red, green and blue sensor elements are stacked in a vertical fashion. In addition, 92% of the 14.1 megapixel Foveon™ F13 CMOS sensor's surface is able to detect light through the use of microlenses that project light onto the stacked red, green and blue pixels.

[00150] In some other embodiments, a conventional CMOS or CCD sensor with Bayer filters can be used. Generally, a Bayer filter-type sensor element consists of one red pixel, two green pixels, and one blue pixel that can detect red, green and blue light simultaneously. However, the optical resolution of a Bayer filter CMOS sensor or a CCD array is typically less than that of a Foveon™ CMOS sensor, where 92% of the surface area is sensitive to red, green and blue light.

[00151] Referring now to FIG. 8B, there is shown an example flow diagram of an image data processing algorithm that may be carried out by a processor in some embodiments to analyze intensity and relative reflectance data.

[00152] Analysis process 840 begins at 845, by illuminating a mineral sample and retrieving image data from the image sensors. The image data can comprise red, green and blue channels.

[00153] At 850, the image data may be analyzed to determine whether any sensor elements (pixels) in the red channel are saturated. If no red pixels are saturated, then the light intensity of the light source (e.g., in all channels) may be increased at 855 and the sample may be re-imaged. Conversely, if greater than approximately 100 red pixels are saturated, then the light intensity may be decreased (not shown). Light intensity can be increased, for example, by increasing the current supplied to a light source (e.g., a diode laser or an LED). If the red light intensity is increased or decreased, then the green and blue light intensities can be increased or decreased by the same proportion.

[00154] At 860, the processor can select photosites that exhibit a change in light intensity with wavelength that is characteristic of gold. For example, photosites where red intensity is at least 45% greater than green intensity, and where green intensity is at least 50% greater than blue intensity can be selected.

[00155] At 865, photosites with a red intensity that is at least 55% of the maximum intensity and with a blue intensity that is no more than 30% of the maximum intensity can be selected from the photosites identified at 860.

[00156] At 870, the RSP can be computed for each of the selected photosites, and the product evaluated at 875.

[00157] If the RSP exceeds 1.75 times the maximum intensity level, the photosite may be identified as a gold candidate at 880. [00158] At 885, the amount by which RSP for a given photosite exceeds 1.75 times the maximum intensity level may be used to calculate how much of the photosite is covered by an image of a gold particle. For example, for an RSP equal to 1.75 times the maximum intensity the fractional gold coverage may be 50%, and for an RSP equal to 2.75 times the maximum intensity the fractional gold coverage may be 100%. For RSP values between 1.75 and 2.75 times the maximum intensity, an interpolation function (e.g., a linear function) can be used to calculate the fractional gold coverage. For RSP values less than .75 times the maximum intensity the fractional gold coverage of a photosite may be considered to be zero, and for RSP values great than 2.75 times the maximum intensity the fractional gold coverage of a photosite may be considered to be 100%.

[00159] At 890, the estimated gold grade can be computed, for example based on equation (8).

[00160] Referring now to FIG. 9, there is shown an example apparatus for detecting gold in accordance with some embodiments. Apparatus 900 comprises at least one light source 905. In one example, at least one light source 905 comprises three LED light sources, producing 420 nm (violet), 525 nm (green) and 660 nm (red) light.

[00161] Light from the at least one light source 905 may be directed to a desired area of the mineral sample (e.g., a sample area approximately 2.5 mm by 4 mm) using, for example, optical fibers 920A and 920B. In practice, a plurality of optical fibers may be used, the outputs of which may be positioned generally equidistantly from each other in a circular pattern surrounding the sample area. FIG. 10 is a plan view illustrating an example placement of optical fibers around a sample area. In FIG. 10, the sample 1050 is illuminated by blue light output from fibers 101 OA, 1010B, 1010C, by green light output from fibers 1020A, 1020B, 1020C, and by red light output from fibers 1030A, 1030B, and 1030C.

[00162] In some embodiments, another illumination method is to arrange the red, green and violet LEDs in a ring that surrounds the lens as shown in FIG. 11. Apparatus 1 100 in FIG. 11 is generally analogous to apparatus 900, except that LEDs 1105 are positioned to replace the emitting tips of the optical fibers 920A and 920B in apparatus 900.

[00163] Referring again to FIG. 9, light reflected by sample 950 can be captured by focusing lens element 930, which can be convex to focus the reflected light onto an image sensor 935. Optionally, a shutter 933 may be used to provide control over the exposure time of the image sensor. Similarly, in apparatus 1 100 of FIG. 11 , light reflected by sample 1 50 can be captured by focusing lens element 1130, which can be convex to focus the reflected light onto an image sensor 1 135. Optionally, a shutter 1133 may be used to provide control over the exposure time of the image sensor.

[00164] In some example embodiments, image sensor 935 (or 1 135) is a 14.1 megapixel Foveon™ F13 CMOS image sensor, focusing lens element 930 (or 1130) is a 65 mm 5x achromatic lens. Focusing lens element 930 may be positioned approximately 28 mm from a cut surface of sample 950 (sample 950 may have been previously cut by a diamond saw).

[00165] Once an image is formed by image sensor 935 (or 1 135), it can be transmitted to a processor 960 (or 1 160) for further processing. Processor 960 or 1160 may be a general purpose processor, a field programmable gate array (FPGA) or other suitable processor, which can be programmed to perform the desired image processing.

[00166] Each of the components of apparatus 900 or 1 100 can be chosen to be small, light and to require low power. Accordingly, apparatus 900 or 1100 can be constructed in a rugged, hand held enclosure, to act as a battery powered instrument, useful in the field for mineral exploration, and for grade control during gold mining. A hand held apparatus generally requires minimal training to use, and would be considerably safer to use than an X-Ray Fluorescence (XRF) detector. In particular, apparatus 900 or 1100 employs only visible light and not high energy ionizing radiation such as X-rays. Apparatus 900 or 1100 can be used safely with minimal precautions other than to avoid directing laser light into a user's eyes, if lasers are employed as light sources. If LEDs are employed then there are few other safety considerations other than to avoid looking directly into the light from the LEDs.

[00167] For ease of exposition, embodiments are described herein primarily with reference to gold detection. However, it will be understood that the methods described herein may also be applicable to the detection of other metals or materials exhibiting a desired optical reflectance characteristic as a function of wavelength. [00168] Some of the embodiments of the systems and methods described herein may be implemented in hardware, or a combination of hardware and software. In particular, aspects of the embodiments comprising software may be implemented in computer programs executing on programmable computers, which may comprise at least one processor, a data storage system (including volatile and non-volatile memory and/or storage elements) and a display. For example and without limitation, the programmable computers may be an embedded device, personal computer, laptop, personal data assistant or mobile device. Program code is applied to input data to perform the functions described herein and generate output information. The output information is applied to one or more output devices, such as a display, in known fashion. Each such computer program is preferably stored on a non-transitory computer readable storage medium.

[00169] Numerous specific details are set forth herein in order to provide a thorough understanding of the example embodiments described herein. However, it will be understood by those of ordinary skill in the art that these embodiments may be practiced without these specific details. In other instances, well-known methods, procedures and components have not been described in detail so as not to obscure the description of the embodiments. Various modifications and variations may be made to these example embodiments without departing from the scope of the invention, which is limited only by the appended claims.