# Method of assessment of air pollution

(57) Abstract:

The method can be used in the field of ecology and refers to remote sensing of the condition index of the atmosphere. The method includes converting the spectral brightness of the image I (x, y) in a digital matrix elements G, R plots of the visible spectrum, logical sorting pixels in both matrices and obtain the resulting matrix R-image of the same size. Next, the calculated numerical characteristics of an electrical signal output matrix, mathematical expectation, variance, spatial envelope of the spectrum. On the basis of the calculated histogram of the distribution of the index of the state of the atmosphere and its autocorrelation function with reference obtained relative distribution of the absolute values of the control sites contained in the original photo. The effectiveness of the method is characterized by efficiency, reliability and statistical stability of the estimates. 4 Il. The present invention relates to the field of ecology, in particular for remote sensing of the condition index of the atmosphere, and may find application in national systems of environmental monitoring.3or MAC,

the average value of one-component impurities in the air and its standard deviation, mg/m^{3}or MAC.

< / BR>

where

- the average annual concentration of i-th substance,

MAC

_{i}- maximum permissible concentration of i-th substance in the ambient air of residential areas,

j is the exponent of iseffectively harmful substances, equal to 0.85; 1; 1,3; 1,5 for substances, respectively, IV, III, II and I hazard classes (see , for example, the Yearbook of the state of air pollution in cities in Russia in 1994, edited by E. Y. Bezuglova, MGO them. A. I. Voeikov, St. Petersburg, 1996).Usually I

_{m}count to five components defining the main contribution to the total pollution of the atmosphere, provided that I

_{1}> I

_{2}> I

_{3}> I

_{4}> I

_{5}.There is a method of assessment index state of the forests, including remote soderzhanie or the degree of destruction of chlorophyll and carotene in the needles, foliage and evaluation by the ratio of chromatic ratios r/g condition index in conventional points (see RF patent N 2038001, class A 01 G 23/00, 1995, similar).The disadvantage of analog is the impossibility of direct use to assess the condition index of the atmosphere.The closest analogue to the technical nature with claimed is a method of estimating pollutant concentrations based on lidar sensing (see , for example, Monitoring of pollution sources. In the book: P. Mieris. Laser remote sensing. Translation from English. M.: Mir, 1987, S. 455 - 460). In the method of the nearest analogue:

probe plot of atmospheric pulse lidar on a controlled route;

take the first channel signal of the primary radiation by elastic backscattering;

take the second channel offset signal inelastic scattering from molecules of the main gas atmosphere N

_{2}or O

_{2};

take the third channel offset signal inelastic scattering from molecules pollutants NO

_{2}or CO

_{2};

normalized signal of Raman scattering from molecules contaminants feedback signal of Raman scattering from molecules jus the Yu backscattering signal receive the pollutant concentration.Disadvantages closest analogue is

the inability to obtain an integrated assessment of the state of the atmosphere in all the ingredients at the same time;

the locality of the results of the evaluations are tied only to track the spread of the beam sensing;

the need for the generation of energy for active sensing of the atmosphere in a selected area of the light range.The problem solved by the claimed method, is to ensure the possibility of rapid remote quantitative evaluation index of the atmosphere of the region and its rate fluctuations on spatial coordinates.This object is achieved in that in the method of valuation of air pollution on the basis of remote receiving a multispectral image of the region, including the control of industrial site, and additionally transform by sampling the analog values of the spectral brightness I (x, y) in a digital matrix G, R images of the visible range of the dimension elements each, with a linear resolution of each element is larger than the fractal plot of the underlying surface, carried out elementwise logical sorting pixels in both matrices in sootwetstwujiut the resulting matrix of the same size, calculate the numerical characteristics of the electrical signal output matrix: mathematical expectation, variance, spatial envelope of the spectrum, autocorrelation function, calculate the histogram of the distribution of pixels for brightness, are binding obtained relative distribution of the absolute values of the condition index of the atmosphere of its values and area control sites.Comparative analysis of the proposed technical solution with the closest analogue shows that the inventive method differs from the known to the introduction of new manufacturing operations, ensuring the achievement of the properties, laws which are manifested in the claimed object for the first time. These new properties, as

statistical stability of the results of the integral evaluation;

documentary reliability assessments, the possibility of zoning the territory of the region on the levels of the MPC and visualization for the consumer;

verification system ground-based measurements, selection of the optimal strategies of local measurements on the ground.This suggests that the proposed method satisfies the criterion of "Inventive step".Therefore, assessment of the condition of the atmosphere in the region on a single measurement is statistically unstable. Due to the high cost and complexity of the method of sampling and subsequent chemical analysis in many ingredients and control points on the ground to obtain timely and reliable assessment is not possible.A complete description of a random variable is a function of a probability distribution. A statistically robust assessment of air pollution should include finding the function p which provides significant advantages. Air pollution in cities directly over industrial sites reaches values of units and even tens of MPC. Physically this is evident in the increasing number of particles of anthropogenic origin in the air, increasing the turbidity of the atmosphere, the formation of smogs. When shooting from space turbidity of the atmosphere determines the interval of change of the ratio of the spectral brightness of the system "atmosphere - underlying surface".The interaction of solar radiation with anthropogenic particles occurs at the molecular level. Complex organic molecules and photochemical origin of the observed broadening of the bands of allowed energy levels. Absorption of molecules photons with high energy (h.), i.e., the short-wave part of the visible range, the re-emission by the excited molecule quanta of energy due to the overlapping bands of allowed energy levels occurs with the smaller values (h

_{i}.). Excited molecules re-emit a series of combination frequencies in the longer-wavelength area of visible range (see, for example, P. Mieris. Laser remote sensing. Translation from English. M. : Mir, 1987, S. 124, PL. 3.4. The wave number of the Raman shift)General components of the visible range, and registered multispectral image anthropogenically contaminated sites have been acquired mainly reddish tint. In addition, in the case of a powerful anthropogenic organic smoke and photochemical origin is more intensive absorption of red and near-IR wavelengths (see, for example, L. I. Chabursky. The reflective properties of natural objects in the range 400 - 2500 nm, 1 o'clock. Min. Defense of the USSR, 1986, S. 103 - 105, S. 134, PL. 10. The coefficients of the spectral brightness aerial haze).Thus, breeding characteristics of anthropogenically contaminated sites in the multispectral space images are orange-red shift of the spectrum of the scattered radiation and dark cherry shade areas of intense absorption. By scanning multispectral image in R-band, receive information on the degree of atmospheric pollution in the form of brightness levels of the matrix elements proportional to the contamination of each plot. However distantly registered the level of reflected solar radiation is both the sum of the reflections from the underlying surface and atmospheric haze. There are methods parametric separation eh formations coefficient spectral brightness of the underlying surface does not exceed 15 - 20%. When turbid atmosphere is slyvania spectral energy and spatial characteristics of the underlying surface. One of the methods of artificial slyvania effect of the underlying surface is scanning the original image with the linear size of a pixel is larger than the fractal plot of this surface (see, for example, the Separation of the effects of the underlying surface and the atmosphere. Final report on the study of the parameters of the system "Atmosphere - surface remote sensing methods. Experiment ISS-M-UF-6 at the station "Salyut-7", 1983 - 1985), Sri Academy of Sciences of the USSR, 1988, S. 23 - 31). Under these conditions, the brightness of the atmospheric haze is averaged over a larger area without distortion numerical characteristics of the desired distribution: mathematical expectation and variance. Effect of recreational abilities vegetation on satellite images with great clarity images at the same time there are green areas with relatively satisfactory state of the atmosphere above them.For highly dynamic atmosphere, on the basis of continuity of a function of condition index for spatial coordinates, to green areas did not distort the resulting statistics, providentialist absorbs up to 93% of the visible radiant energy. Most healthy forested areas on multispectral images have a dark green color. Therefore, the dark green areas G-band mirror the environmentally unfriendly, intensely absorbing radiant energy of smogs R-band. Based on the connection between chromatic coefficients (see analog, RF patent N 2038001, 1995) logical sorting of the elements of the matrices G and R are carried out in accordance with the algorithm: if R (G, R, and if R < G, where k is the coefficient of correlation between chromatic coefficients (r, g).The algorithm is implemented by the following program:

< / BR>

After sorting receive the resulting matrix and calculate its numerical characteristics: mathematical expectation, variance. Well-known mathematical procedures (see, for example, calculating the statistics of a random sample. In kN. : Handbook of mathematics. G. Korn, T. Korn. Translation from English. M.: Nauka, 1979, S. 537 - 539) receive a histogram of the sampling distribution of the probabilities of the level of brightness of the resulting R-image. In Fig. 1 shows one implementation of the sampling distribution, normalized relative standard deviation (I/). Binding obtained otnositeljnoyj sites. The pairing process described in example implementations of the proposed method.Another, the most informative parameter of the monitored process, along with the probability distribution function of the integral parameter is the speed fluctuations of the index on the spatial coordinates. This parameter carries information about stagnation in the atmosphere over the city or Vice versa about her "ventilation". A quantitative measure of the speed fluctuations of the random processes are their autocorrelation function. By definition (see , for example, Driveways A. M. fundamentals of statistical radio engineering. M : Communications-Izdat, 1964, S. 94, formula 7.35) the autocorrelation function of B(L) process is calculated as the inverse Fourier transform of its power spectrum S(F)

< / BR>

In connection with the development and practical application of the algorithm of fast Fourier transform (FFT) the whole procedure of calculation of the autocorrelation function of the index of the state of the atmosphere can be automated. For this purpose, we first calculated the two-dimensional spatial spectrum G (f

_{x}f

_{y}) the output matrix software FFT calculation in accordance with addiction

< / BR>

This operation come in the hands of the Earth, GENASYS, Inc., San Diego, USA).By integrating the two-dimensional spatial spectrum on the ring segments, in accordance with the programmatic calculation (see, for example, Methods and apparatus for digital image processing, version 2.1, 2.2. Mask of integration of the Fourier spectrum, technical report, MSU, 1986, S. 61 - 64) allocated spatial envelope of the spectrum G (1/L) as a function of the spatial period (L). The energy spectrum of the signal S (F) is related to its amplitude spectrum of G (1/L) ratio

< / BR>

where

AB is the interval of definition of the function G (1/L), in this case, the linear dimensions of the picture region (see , for example, Driveways A. M. fundamentals of statistical radio engineering. Communication-Izdat, M., 1969, S. 93, formula 7.30).Since the energy spectrum of a signal is an even function of frequency (negative frequencies cannot be), then the autocorrelation function is calculated as the cosine transformation from the energy spectrum

< / BR>

The last integral is calculated by the software calculation (see, for example, B. N. deacons. Reference "System MATH CAD, Radio and communication". M., 1993, S. 61).An example implementation of the method.The inventive method can be implemented on the functional diagram of Fig. 2. The technical structure is engaged in the survey area (from 50 x 50 to 200 x 200 km) of interest 3, containing a key industrial area 4. Switching the camera 2 over predetermined areas of observation 3 is carried on-Board control system 5 according to the programs transferred from the mission control Center 6 via the radio channel control 7. Images with the environmental conditions of the studied areas quickly in the composition of the shot capsules 8 are delivered to the PMU. After processing the captured frames arrive at the information processing Center 9. Each image region 10 sequentially scanning through the input device information 11 (type Panasonik) in the form of digital files (matrix m x n) R and G image is recorded in the random access memory of the processor 12. The set of programs specialized digital processing of images ER MAPPER 5.0, MATH KAD, is recorded on the HDD 13. Processed arrays of information (resulting matrices) in the form of histograms of the distributions, Fourier spectra, autocorrelation functions are visualized on the display 14 with printing and document the results of calculations on a color printer 15 (type UGA). As the PC image processing use stations like SUN or IBM PC 486/487.In Fig. 1 presents a histogram of the distribution of Indiana. Numeric characteristics of the processed image area were as follows: amount entered in the PC image area 4.3 x 4.3 cm, scale snapshot 1 cm : 1,7 km, the Size of the scanned matrix 675 x 675 elements, the spatial resolution is 10.8 meters per pixel. The range of values of the quantized image brightness levels 0 - 256 mathematical expectation of an electrical signal matrix reached the amount of 203.9 and a standard deviation of 61.Since changing the shooting conditions (seasons, days, azimuth, altitude of the Sun changes the brightness of the image, and the histogram is built in the brightness values, normalized relative standard deviation (I/). Under these conditions, the change in the average brightness and contrast of the image does not affect the distribution of the resulting statistics. It is known (see, for example, theoretical foundations of radar. Edited Century. Dulevich. M : Owls. radio, 1964, S. 114) that a random variable, distributed in the interval from zero to sporadically maximum values, subject to the generalized Rayleigh law. At a known relative law of the resulting distribution of brightness levels involved in its binding to the absolute index values comprising the x industrial sites (C-d Khimmash, Asterisk and others), where MPC > 5, amounted to S - total area of the city. By definition, the probability of true expression

< / BR>

From the graph of Fig. 1 find that the square of the "tail" of the curve is equal to 0.1, the ratio whence = 2,2 MAC. The absolute scale of values of the condition index of the atmosphere, Sergiev Posad provided coaxially with a relative scale (I) in Fig. 1. In Fig. 3 shows a graph of the spatial spectrum of the condition index of the atmosphere, and the graph of Fig. 4 - its autocorrelation function, obtained by the method of software calculation of the above ratios. From the graph of Fig. 4 it follows that the interval of the correlation corresponds to 0.75 km, i.e., in the city there are areas with a length of approximately 1.5 km where there is stagnation, "hanging" can be customized. More than 1.5 km in the city do not host point ground-based measurements of control of the air environment, because such samples are not independent from each other and do not carry additional information, and lead to unnecessary costs.The effectiveness of the proposed method is characterized by parameters such as efficiency, reliability, statistical stability, cost. The cost of one measurement and chemical analysis FOSS city the number of dimensions NO

_{2}, CO

_{2}, SO

_{2}up to several thousand per year. The total costs for the ground-based measurements of the city sanitary-epidemiological supervision is estimated at hundreds of millions of rubles. In the claimed method in the example implementation, the volume of a sample of measurements is 675 675 = 455625 elements-counts, which provides statistical stability and reliability of assessment results. The cost of one multispectral image by world prices: 400 - $ 600, the processing cost is $ 200, the cost of processing tools $ 15,000. Thus, the cost of implementing the proposed method are repaid within one year. Method of assessment of air pollution on the basis of remote receiving a multispectral image of the region, including the control of the industrial area, characterized in that it further transform by sampling the analog values of the spectral brightness I(x, y) in a digital matrix G, R images of the visible range of the dimension elements each, with a linear resolution of each element is larger than the fractal plot of the underlying surface, carried out elementwise logical sorting pixels in both matrices in accordance with the algorithm, if R (G, trico the same size, calculate the numerical characteristics of the electrical signal of the output matrix is the mathematical expectation, variance, spatial envelope of the spectrum, autocorrelation function, calculate the histogram of the distribution of pixels for brightness, are binding obtained relative distribution of the absolute values of the condition index of the atmosphere in the region of its values and area control sites.

**Same patents:**

The spline interpolator // 2116669

The invention relates to automatic control and computer engineering

The spline interpolator // 2116668

The invention relates to automatic control and computer engineering

Device for digital filtering // 2055396

The invention relates to computer technology, in particular to the processing of information, and can be used in devices receiving the information transmitted by the communication channel, in which there is a possibility of distortion, i.e

The invention relates to computer technology and can be used in specialized high-performance computing systems for solving least squares, solving systems of linear algebraic equations for the case when the number of equations is greater than the number of unknown or no confidence in a good condition matrix

Device for interpolation // 2042979

The invention relates to computer technology and can be used to perform functional transformations in a variety of digital information processing systems, such as onboard aircraft systems

Digital filter // 2029362

The invention relates to digital signal processing and can be used to filter the valid signals in radio devices for various purposes, in particular in radar

Multichannel digital filter // 2029361

The invention relates to computing, and in particular to means of digital signal processing in real time, and can be used in radar, navigation and communication for construction of high-speed digital processors

Linear interpolator // 2028665

The invention relates to the processing and transmission of signals, where the linear interpolation of signals at high speed, when working in real time, such as telemetry, automation, communications, computing, etc

The device for calculating the convolution // 2112274

The invention relates to the field of computer engineering and can be used in specialized computer systems to calculate swerski

Matrix setprocessor // 2079879

The invention relates to computing and is designed to build on the basis of specialized computers

The invention relates to computer technology and can be used in specialized computer systems to calculate the two-dimensional convolution

Device for multiplying two matrices // 2049352

The invention relates to computer technology and can be used in specialized computing systems, in particular, in digital signal processing for multiplying two matrices

The invention relates to computer technology and can be used in the automated command and control system for managing the movement of different types of vehicles on the road network with different cross-sections

Vector accelerator // 2042980

The invention relates to computer technology and can be used in the production of vector-based devices, providing high performance vector operations

The invention relates to computer technology and can be used for performance analysis of Queuing systems

A device for processing matrices // 2037200

The invention relates to computing and microelectronics, and can be used in specialized pre-processing of multidimensional signals