BACKGROUND OFTHE INVENTION

Particulate matter or PM is an air pollutant consisting of a mixture of particles that can be solid, liquid or both are suspended in the air and represent a complex mixture of organic and inorganic substances. These particles vary in size, composition and origin. Numerous scientific studies have indicated that the most harmful component in air pollution is the microscopic dust with diameter less than 10 micrometers (PM10). This is due to their ability to penetrate deep into the lungs and embeds itself in the lungs. This effect has been linked to respiratory disease, cancer and other potentially deadly illnesses (Pope et al, 1995; Kenneth Donaldson et al, 2000). For preventing long exposure to this type of harmful air pollution that causes adverse health effects, it motivates a growing interest to develop efficient techniques to monitor this type of pollution (Bruzzone & Prieto, 2002). For instance, U.S. Pat. No. 4,921 ,349 discloses a photographic method of collecting and reducing photographic film density data for monitoring air quality is disclosed in which photographic film is first calibrated using calibration scales of known reflectance and the measured radiance of one calibration scale step to relate the photographic film density to the known radiance. Measurement of a target's radiance and sight path radiance is made via target photographic image density analysis and mathematical expressions which convert density information into radiance and/or atmospheric transmittance data using expressions developed from the calibration scale data.

Therefore, there is a need for a simple, reliable air pollution monitoring method which used equipment of reasonable cost and which can be easily used in remote location. Moreover, there is a need to monitor real time PM10 air pollution at multi-location. This is an attempt to fulfill the need for preventing long exposure of this harmful air pollution.

SUMMARY OF INVENTION

Accordingly, the present invention provides a method to obtain quantitative data for determining particulate air quality concentration level of a sight path towards a target which is located more than 20 meters from an image sensor including the steps of obtaining an image from the image sensor focusing on a particular target and measuring solar irradiance by using an atmospheric radiation measurement device, separating the image into red, green and blue bands, extracting average digital number for red, green and blue bands from the digital image, converting the average digital number for red, green and blue bands to irradiance and calculate reflectance values, obtaining an atmospheric reflectance from a referenced reflectance and the reflectance values obtained from the image, determining the relationship of atmospheric reflectance and PM concentration level by our own developed algorithm as shown in equation (4), thus obtain the algorithm to determine PM10 concentration and producing the particulate air quality concentration level from this algorithm.

Furthermore the present invention also provides a system for monitoring particulate matter air pollution concentration level in an environment, including an image sensor adapted to capture an image which is located more than 20 meters from the image sensor through atmosphere suspended particulate matter, an atmospheric radiation measurement device for measuring solar radiation and at least one computation means having a PM concentration measurement program configured to measure a particulate concentration level from an image captured by the image sensor. Also provided is a method to obtain quantitative data for determining particulate air quality concentration level of a sight path towards a target which is located less than 20 meters from an image sensor including the steps of obtaining an image from the image sensor focusing on a particular target, separating the image into red, green and blue bands, extracting average digital number for red, green and blue bands from the digital image, converting the average digital number for red, green and blue bands to irradiance and calculate reflectance values, obtaining an atmospheric reflectance from a referenced reflectance and the reflectance values obtained from the image, determining the relationship of atmospheric reflectance and PM concentration level by our own developed algorithm as shown in equation, thus obtain the algorithm to determine PM10 concentration and producing the particulate air quality concentration level from this algorithm.

Finally, the present invention also provides a system for monitoring particulate matter air pollution concentration level in an environment, including an image sensor adapted to capture an image which is located less than 20 meters from the image sensor through atmosphere suspended particulate matter and at least one computation means having a PM concentration measurement program configured to measure a particulate concentration level from an image captured by the image sensor.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is an air pollution measuring and warning system in accordance with an embodiment of the present invention.

FIG. 2 is an arrangement of the system of FIG. 1 to obtain a reflectance value of a reference target. This arrangement also used to calibrate the digital camera.

FIG. 3 is an arrangement of the system for method 2.

DESCRIPTION OF THE PREFERRED EMBODIMENT

The present invention relates to an air pollution measuring and warning system (100) for monitoring particulate matter air pollution or PM concentration level in an environment. Thereof, the system (100) can be configured with a predetermined PM concentration threshold value whereby once the threshold value has been exceeded, the system (100) triggers its alert and notification procedures. Alerts can be relayed over a variety of communication protocols including e-mail, encrypted internet or directly into linked information systems. In addition, the system (100) may utilize existing network surveillance video camera, which will allow additional features of automatic and real-time

PM concentration monitoring at multi-location.

The system (100) monitors air pollution by accurately measure PM concentration level. Moreover, such measurement can be done continuously by the system (100). The measurement of this system (100) was compared to the measurement of DustTrak™ meter. This comparison results have shown to have a high correlation coefficient (R ² ) and low root-mean-square error (RMS), this indicates that the measurement of this system has a high degree of accuracy. Referring to FIG. 1, the system (100) generally comprises a camera (110), an atmospheric radiation measurement device (120) and at least one computation means (130) that includes a PM concentration measurement program.

Suitably, the camera (110) is a digital camera and it is connected through either a wired or wireless connection to the computation means (130). An analog camera may also be utilized by the system (100). This is done by connecting the analog camera to an analog- to-digital converter means and connecting the analog-to-digital converter means to the computation means (130) through either a wireless or wired connection. The camera (110) is adapted to capture image through atmosphere suspended particulate matter (150). The image that includes PM concentration is analyzed by the computation means (130).

The atmospheric radiation measurement device (120) is suitably a spectroradiometer to measure the solar radiation. The RGB information extracted from the device (120) is used by the PM concentration measurement program as dependent variables in deriving the air quality information or PM concentration level from the image captured by the camera (110).

The computation means (130) is adapted with the PM concentration measurement program to analyze the image from the camera (110) and solar radiation information from the radiation measurement device (120) in order to determine the PM concentration level. Such computation means (130) includes computer, server, personal digital assistant and etc. If the PM concentration level is more than a predetermined threshold value, the computation means (130) triggers an alert over a variety of communication protocols including e-mail, encrypted internet or directly into linked information systems or any other alert devices for notification purpose of the air pollution level. However, if the PM concentration level is below the predetermined threshold value, the computation means (130) stores the concentration information in a storage medium such as hard disk, CD, DVD and the like for recording purpose. The PM concentration level measurement program utilizes RGB information measured by the atmospheric radiation measurement device as calibration factors on atmospheric radiation, and then estimating the atmospheric reflectance from the camera (110) data to generate unique image processing algorithm. Thereon, the image processing algorithm is used to obtain the PM concentration level from the image captured by the camera (110). The algorithm is based on atmospheric aerosol characteristic model and this in turn can be related to the air pollutants or PM concentration.

Moreover, the system (100) is able to obtain quantitative data for determining particulate air quality concentration level of a sight path towards a target (160) from a digital image by using an algorithm of the PM concentration level measurement program. The algorithm is developed by using a method that comprises the steps of: (i) extracting average digital number (DN) for red, green and blue (RGB) bands from the digital image and simultaneously, collecting solar irradiances by using the atmospheric radiation measurement device (120); (ii) converting the average DNs to irradiance and reflectance values; and (iii) determining co-relationship of atmospheric reflectance and PM concentration level.

The method to obtain quantitative data for determining air quality concentration level will be described by way of the following example, which does not limit the scope of the present invention.

Example 1

The system is used to measure the PM concentration level for particulate matter less than 10 micron (PM10).

A digital camera (100) is used as an image sensor to capture digital images of a referenced target (160). Thereon, the digital images are separated into three bands namely red, green and blue bands for multispectral algorithm calibration. In-situ measurements of corresponding air pollution parameters are measured. These in-situ measurements are then used as the dependent variables in deriving the air quality information using the digital camera data.

The digital numbers of the three bands are converted into irradiance and then reflectance. The relationship between the reflectance and the corresponding air quality data is determined using regression analysis. A new algorithm is developed for detecting air pollution from the digital camera images chosen based on the highest correlation coefficient, R and lowest root mean square error, RMS for PM10. The algorithm used is based on the apparent reflectance values detected at near and far distances from the reference surface, and these in turn can be related to the concentration of the air pollutants. The coefficients of the calibrated algorithm are determined and used in estimating the air pollution level. Comparison was made between the air pollution parameter and the results using different colour paper or wall of a building as a reference surface. It was found that the red colour paper produced the best result when it was used as a reference surface in this study. The key issue in retrieving atmospheric reflectance from remotely sensed data is to identify the surface reflectance. The reflectance measured from the sensor (reflectance at the top of atmospheric) is subtracted by the amount given by the surface reflectance to obtain the atmospheric reflectance. And then the atmospheric reflectance can be related to the PM10 by using the regression algorithm analysis.

Digital images were captured using the digital camera (110) with the camera axis was at 90° with the plane of the reference target (160). The digital images of the reference targets (150) were captured by using the digital camera at near and far distances (160) from the targets (150) as shown in FIG. 2. The particulate matter air pollution was in between the reference target (150) and the digital camera (110). The spectral reflectance values of the four colour papers were measured using a handheld spectroradiometer. The spectroradiometer also used to measure the atmospheric radiation for multi-spectral analysis.

The digital imageries were separated into three visible wavelength bands namely red, green and blue bands. The DN values were converted into irradiance using the digital camera coefficient calibrated for each band using equation (1). The digital camera (110) observed signal is the sum of the surface reflectance and atmospheric reflectance.

Y= a X+ b (1) where:

Y = irradiance for one of the RGB band (Wm ^"2 nm ^"1 );

X = digital number for one of the RGB band;

a, b = constant (determined from calibration). The key issue in retrieving atmospheric reflectance from remotely sensed data is to identify the reference reflectance. After that, the reflectance recorded by the digital camera (110) was subtracted by the reflectance of the known surface to obtain the reflectance caused by the atmospheric components.

There are three techniques used in this study which are semi empirical, far-near and normalization flat-field techniques. For a first technique, over a simple black target, the observed atmospheric reflectance is the sum of reflectance of aerosols and Rayleigh contributions. This simplification, however, is not valid at short wavelengths (less than 0.45 pm) or large sun and view zenith angles (Vermote and Roger, 1996).

The atmospheric reflectance can be obtained by subtracting the reference reflectance from the total reflectance observed by the digital camera (110). Then, the retrieved atmospheric reflectance values are used for regression analysis to determine the relationship between atmospheric reflectance and PM10 concentration.

For the second technique, the digital images of the reference targets were captured by using the digital camera (110) at near and far distances from the targets (150). The far distance between a building as a reference target (150) and the camera (110) was 100 meter. Presumption made in this study was that the air quality measurement represents a column of 100 x 5 x 5 meter ³ around the air pollution station.

Third technique is a normalization technique called flat-field calibration. This technique has been used in other studies (Roberts et al., 1986, Carrere and Abrams, 1988, and Murphy et al., 2004). The model algorithm is shown in equation 3. This technique normalizes the DNs in each image to relative reflectance. It assumes that there is an area in the scene that is spectrally neutral (no variation in reflectance with wavelength). For each band in each image, the raw DNs values over the four colour papers are extracted and averaged. The DNs values for the rest of the four colour papers are converted to relative reflectance using equation 3 below. The data are used in the multiple regression algorithms. The relative reflectance value for each image is computed using equation (3).

where:

p = relative reflectance;

p _r = reference reflectance,

p _r for red colour paper = p _rRR , p _rRG and p _rRB for red, green and blue bands respectively,

P _r for green colour paper = P _Γ GR, P _Γ GG and P _Γ GB for red, green and blue bands respectively, p _r for blue colour paper = prB _R , pmo and PΓBB for red, green and blue bands respectively, p _r for black colour paper = p _rDR , P _Γ DG and p _rD β for red, green and blue bands respectively);

DN, = digital number for image of each band (Wall);

DN _r = digital number for reference target of each band.

The relationship between the atmospheric reflectance and the corresponding air quality data is determined using regression analysis. The coefficients of the calibrated algorithm are determined and used for estimating the air pollution level. Thus, the equation below is used to determine the PM10 concentration level:

P = a ₀ Ra _t m (λ i) + a , Ro _n , (A ₂ ) + a ₂ R ₃ Im (λ ₃ ) + a ₃ (4) where:

P = PM10 concentration level;

R _atm (A ₁ ) = atmospheric reflectance, 1 = 1 , 2 and 3 are the band number of RGB;

a,= algorithm coefficients, j = 0, 1 , 2 and 3 are then empirically determined.

This algorithm can be used to compute PM10 concentration in indoor and outdoor.

Advantages of the Air Pollution Measuring and Warning System

Automated and remotely accessible features of the system (100), enables it to offer low implementation, operation and maintenance cost of ownership. Moreover, it can be easily integrated into personal computer to meet individual requirements for alert thresholds and notification procedures.

In addition, with the capability of the system (100) to continuously monitor air pollution level, air pollution level can be identified rapidly when pollution levels are approaching environmental department defined thresholds for warning or alert conditions. Applications of the Air Pollution Measuring and Warning System

The system (100) can be applied in various ways such as but not limited to:

a. alerting public to prevent long exposure to polluted air;

b. identifying source of air pollution and taking necessary action;

c. addressing the air pollution of a geographic area for tourism information; d. survey information; and e. assessing a particular area for industrial development or any other development by determining the air pollution level of that particular area.

Moreover, the system (100) can be used as a warning alarm or as a monitoring device to complement existing environmental air quality control process.

Although described in the embodiments that the camera (110) is used to capture image for determining the PM concentration level, it is appreciated by a person skilled in the art that the camera (110) can be substituted by an image sensor or any other device that is capable in capturing images.

While embodiments of the invention have been illustrated and described, it is not intended that these embodiments illustrated and describe all possible forms of the invention. Rather, the words used in the specifications are words of description rather than limitation and various changes may be made without departing from the scope of the invention.