BACKGROUND [0002] Determining the amount of impurities in products being manufactured is important especially in the industry for making pulp sheets, paper, and board. Dirt entering a sample during manufacturing in particular constitutes the impurities. The determination of the quantity and quality of im- purities can be made either visually or mechanically. In visual determination, a sheet-like pulp, paper, or board sample is placed on top of glass that supports the sample on a light desk in such a manner that the light from the lamp inside the light desk shines through the glass and sample. Alternatively, or in addition to this, the sample can be lighted from the top depending on the measurement.

The examiner of the sample looks at the sample from the top, from the direc- tion of its surface normal and searches for macroscopic impurities and classi- fies them by size and possibly by the tone of colour. The sample can be meas- ured dry, or the sample can also be doused, so as to more easily detect dirt under the surface. In the determination, the sample can be examined sepa- rately from both sides.

[0003] Instead of visual examination, a camera can be directed to the sample to form an image of the sample and the dirt in the sample. In such a case, a dried pulp, paper, or board sheet sample is picked from a finished pulp bale, set to lie on a table, and fastened to a pulling device, with which the sample can be horizontally pulled along the glass surface of the table. In this case, too, the sample can be measured either dry or wet. Above the sample, there is a camera that moves transverse to the pulling direction of the sample and scans image tracks from the sample for analysis with an image-processing program and consequently, for determining by program the sample properties related to the amount and possibly also quality of dirt in the sample. Impurities' measurements of this kind are standardized in standards ISO 5350-1, ISO 5350-2, T 213 om-97 and T 437 om-96, for instance.

[0004] There are, however, problems related to these measure- ments. Visual examination is very difficult, slow, and inaccurate, and the result depends on the examiner. In both the visual and camera examination, dust

and dirt accumulate on the glass of the measuring table and cause errors in the measurement, which is why the glass supporting the sample needs to be continuously cleaned. Cleaning is, however, difficult, and the dust and dirt can never be completely removed from the glass. In addition, the dousing of the sample sheet is very exact, because water cannot easily exit the table surface and causes errors in the measurement. Turning the wet sample sheet to measure the other side may break the sheet and can, thus, in many cases be completely impossible. In camera measurement, filming is only easy to arrange on one side of the sheet (the top side), and dirt on the other side of the sheet remains undetected. However, if the filming is done on both sides of the sam- ple sheet by using a camera inside the table, there is always more water un- derneath the sample sheet laying on the table, and the image and measuring result are thus different than when taken from above. In addition, the camera must then be protected from water and the protection must be tight. This adds to the complexity and price of the measuring arrangement. Glass sheets or other transparent covers also get dirty, which causes errors in the measure- ments as well as maintenance problems.

BRIEF DESCRIPTION [0005] It is an object of the invention to implement an improved method for measuring impurities, and an arrangement implementing the method, in which dirt accumulation is minimal and filming is simple.

[0006] This is achieved by a method for automatically measuring impurities in a sample. The method comprises lifting the sample upward in such a manner that of the sample, at least the part to be measured is sus- pended without support; filming the part of the sample to be measured with at least one camera from a direction other than the vertical direction; and measur- ing by means of automatic image processing the impurities in the sample.

[0007] The invention also relates to an arrangement for automati- cally measuring impurities in a sample. The arrangement comprises means for moving the sample, and the means for moving the sample are arranged to lift the sample in such a manner that of the sample, at least the part to be meas- ured is suspended without support; at least one camera for filming the part of the sample to be measured from a direction other than the vertical direction; and an automatic image-processing device that is functionally connected to at least one camera, and the automatic image-processing device is arranged to

measure the impurities in the sample in at least one image formed by at least one camera.

[0008] Preferred embodiments of the invention are disclosed in the dependent claims.

[0009] The invention is based on the idea that a sheet-like sample is lifted upward in such a manner that of the sample, at least the part to be measured is suspended freely without touching anything. The sample can then be measured from either one or more directions.

[0010] The method and system of the invention provide several ad- vantages. The sample is not in contact with a support surface that causes er- rors in the measurement, and the measuring result of the sample impurities is not dependent on the measuring direction.

LIST OF FIGURES [0011] The invention will now be described in greater detail by means of preferred embodiments and with reference to the attached drawings, in which Figure 1 shows the measurement of sample impurities, Figure 2 shows the principle of scanning, and Figure 3 shows two-sided sample measurement.

DESCRIPTION OF THE EMBODIMENTS [0012] The presented solution is suited for measuring impurities in different sheets and sheet-like pieces without being limited to them, however.

Possible samples to be measured include pulp sheets, paper, board, fabrics, wood panels, chipboards, hardboards, plywood boards, plastic boards, and glass sheets.

[0013] Let us first examine the presented solution by means of Fig- ure 1. The sample 100 is not on the board, but lifted in the upright position from one edge by means of sample moving means 102, which have mechanical suspenders, such as clamps or pressing jaws. Of the sample 100, at least the part to be measured, which is either the entire sample 100 or a part of it, is then suspended without support. The sample may hang freely in the upright position and only be in contact with the sample moving means 102, but it is also possible that the bottom part of the sample 100 is in contact with a struc- ture, for instance a shredder (the part of the sample that is in contact is, how- ever, then not measured). In such a case, the sample 100 should, however,

not be unduly pressed or stretched so as not to damage it. Cameras 104 and 106 form an image of the sample 100 on different surfaces 1000,1002 in the filming area 108,110 seen by the cameras, and the cameras are directed to the sample 100 in a direction other than the vertical direction. The cameras can be directed to the surface normal of the sample or in some other direction.

The filming direction of the cameras 104,106 can be horizontal, if the sample 100 is upright. The filming is done using optical radiation, and radiation with a wavelength range of 40 nm to 1 mm is considered optical radiation. The filming can be done with monochromatic radiation or by using a desired optical radia- tion band. Many applications use the band of visible light (approximately 400 nm to 700 nm). The filming areas 108,110 and fields of vision of different cameras can be the same or vary from each other, in which case the measur- ing areas 108, 110 on different surfaces 1000,1002 can be equal or different in size. Differing from Figure 1, it is also possible to use only one camera in the measurement, or more than two cameras that are on the same side or on dif- ferent sides of the sample 100. The cameras can have one or more detecting elements, on which the real image of the sample 100 is formed. When using cameras with a pixel matrix as the detector, the entire sample part to be meas- ured can be filmed at one time. Instead of this, it is also possible to scan the part to be measured and to film different points of the part to be measured at different times. During scanning, either the camera or the sample moves.

Scanning the part to be measured is especially necessary when using a cam- era with a pixel line as the detector or only one detecting pixel or element.

[0014] The images formed by the cameras 104,106 are transmitted to an image-processing device 112 that measures the impurities of the sample 100 by means of automatic image processing. The cameras 104,106 can pro- duce digital images or the image-processing device 112 can convert the im- ages of the cameras to digital ones, after which the image-processing device 112 processes the images digitally. The image-processing program of the im- age-processing device 112 can easily compare images formed on different sides of the sample 100 regardless of how the images are formed with the cameras. The image-processing device 112 can determine the dirt count of the sample, the total surface area of dirt, the total surface area of dirt in relation to the measured surface area, the distribution of dirt in the sample, the shape of dirt and distribution according to shape, the colour of dirt and distribution ac-

cording to colour, etc. If colour is not of interest, the cameras 104,106 can also be black-and-white cameras instead of colour cameras.

[0015] The sample 100 can be measured dry or wet. A dry sample does not necessarily be transparent or translucent, but the measurement can also merely determine the impurities on the surface of the sample. The pur- pose of dousing is to increase the transparency of the at least partly transpar- ent sample. Dousing can be done by dousing devices 114 and 116 that spray dousing agent on the sample 100, whereby the transparency or translucency of the doused part 118 improves and the dirt inside the sample 100 are distin- guished better. The dousing agent can be water, but instead of water, it is pos- sible to use different greases, oils or other agents that improve the transpar- ency of the sample 100. According to Figure 1, the cameras 104,106 can be similarly placed with respect to each other on different sides of the sample 100.

The flowing of the dousing agent does not harm the cameras 104,106, be- cause the dousing agent flows down along the sample. When the sample is upright, about the same amount of dousing agent is absorbed in the sample 100 when the excess dousing agent flows away, and dousing affects the opti- cal properties of the sample 100 in the same way on both sides of the sample.

[0016] The dousing of the sample 100 can be done slightly higher than the cameras. The sample 100 is then lowered with the sample moving means 102 downward so that the doused area 118 enters the filming areas 108, 110 of the cameras 104,106. The undoused top part of the sample 100 thus remains dry, which reduces the possibility that the weight of the dousing agent absorbed in the sample 100 cuts the sample 100 that is weakened by dousing at the point being measured or above it (where it has not yet been measured). Dousing should especially be avoided or delayed until the end of the measurement in the area where the suspenders 102 hold the sample 100.

When the sheet-like sample 100 is in the upright position, loose dirt, such as dust, does not as easily collect on the surface of the sample 100 or the optics of the cameras or on the back surfaces of the sample, and the surfaces are easier to keep clean than on a horizontal plane. In the presented solution, the entire measurement can be made automatic, whereby sampling, measuring and sample processing after measuring is done without human action.

[0017] Let us now examine by means of Figure 2 an example on how the sample 100 can be filmed. The camera films transverse tracks 200 to 206 of desired length horizontally of the sample. The sample 100 is moved

after the filming of each transverse track 100 to 206 downward by the length of the track so that the entire surface area gets filmed. Thus, transverse track 200 is filmed first and then transverse track 202. In the situation of Figure 2, trans- verse track 204 is being filmed, and transverse track 206 is next. Each trans- verse track can be filmed at one time, or as shown in Figure 2, each transverse track can be filmed in five different filming sets when the camera scans hori- zontally across the sample 100. If the transverse track is not filmed at one time, it can generally be filmed in two or more sets. Image quality, the number of pixels on the detecting surface of the camera, and the like affect the number of filming times. For instance, the transverse track 204 being measured is filmed in such a manner that image field 2040 is filmed first and after this, the camera moves to the right and films image field 2042. From image field 2042, the camera moves to film image field 2044, and this way, the camera proceeds to image field 2046 and finally to image field 2048. All other transverse tracks are filmed in the same manner. The sample 100 can thus be filmed by scan- ning the sample both in the horizontal and vertical direction. Each image field is formed with one or more pixels of the camera.

[0018] Figure 3 shows one way of positioning the cameras 104,106 and the lighting of the sample 100. In this solution, the sample 100 is translu- cent and the measurement is done by transillumination. In this solution, the camera 104 and an optic radiation source 300 are arranged to remain aligned on different sides of the sample 100 during the entire measurement. Similarly, the camera 106 and an optic radiation source 302 are arranged to remain aligned on different sides of the sample 100 during the entire measurement.

The camera 104 and optic radiation source 300 can be kept aligned during the entire measurement by connecting the camera 104 and the optic radiation source 300 to each other with fixed structures. Alternatively, the camera 104 and optic radiation source 300 can be controlled to move in the same way. The same can also be done to the camera 106 and optic radiation source 302. By placing the pair formed by the camera 104 and optic radiation source 300 at a different height than the pair formed by the camera 106 and optic radiation source 302, it is possible to film with both cameras 104 and 106 at the same time. When the pairs are at the same height, filming should be done at different times. The optic radiation source 302 can then be switched off when the cam- era 104 films and the radiation source 300 illuminates the sample. Correspond- ingly, the optic radiation source 300 can be switched off when the camera 106

films and the radiation source 302 illuminates the sample. When the sample 100 is filmed, the cameras 104,106 can scan the sample both in the vertical and horizontal direction.

[0019] The positions of the camera 104 and optic radiation source 302 in relation to each other can also be kept unchanged. Then the positions of the camera 106 and optic radiation source 300 in relation to each other also remain unchanged. This way, the combinations of the camera 104 and optic radiation source 302 and the camera 106 and optic radiation source 300 can scan the sample 100 by moving in the same way. The optic radiation sources 300,302 direct optic radiation to the sample 100, and the cameras 104,106 form an image of the sample 100 by using the radiation. The optic radiation sources 300,302 can be optically broadband or narrowband, continuous or pulsed (stroboscope), and they can be lasers, diodes, filament lamps or gas- discharge lamps, etc. The images formed by the cameras 104,106 are fed to the image-processing device 112 that measures the impurities of the sample 100 by means of the images. Images taken from different sides of the sample 100 at locations corresponding to each other can be compared with each other in the image-processing device 112, and each individual dirt that shows on both sides of the sample, can be noted as one piece.

[0020] After the measurement, the pulp, paper, or board sample in particular can be returned to stock. A pulp sheet can for instance be dropped down to a water basin, in which the pulping and sewage of the sample is ar- ranged, or a belt conveyor with which the sample can be conveyed to a pulper or back on top of a pulp bale. The sample can also be shredded after the transverse track has been completely filmed.

[0021] The presented solution can also be part of a bigger measur- ing system, in which the sample is measured automatically for numerous dif- ferent properties. The measuring can then be done automatically in a process producing a sheet-like product. A measurement performed according to the presented solution can be made to correspond to a standard measurement.

[0022] Even though the invention has been explained in the above with reference to examples in accordance with the accompanying drawings, it is apparent that the invention is not restricted to them but can be modified in many ways within the scope of the inventive idea disclosed in the attached claims.