A smog pollution source tracing system and source tracing method based on video monitoring

Abstract

The invention discloses a smog pollution source tracing system based on video monitoring and a smog pollution source tracing method thereof. Firstly, a plurality of cameras on a lens group are used to respectively collect street information in different time periods at a plurality of locations, and the collected video information is clipped to generate pictures at the same time point at different locations; then, a digital image method is used to obtain pixel point data information of the pictures, and gradient processing is performed to obtain smog pollution parameters; the smog pollution parameters at different locations are applied with the Poisson equation to obtain the smog pollution concentration at a certain time point at each location; the smog pollution parameters at different time points at the same location are derived to obtain the change value of the smog pollution parameter; then, the value after the second-order derivative of the smog pollution concentration and the change value of the smog pollution parameter are fitted to obtain that the picture shooting position represented when both values ​​are maximum is the smog pollution source. The data obtained by the invention are more timely and can achieve the effect of accurate smog control.

Description

Technical Field

[0001] The invention belongs to the technical field of haze pollution monitoring, and in particular relates to a haze pollution source tracing system based on video monitoring, and also relates to a source tracing method of the haze pollution source tracing system. Background Art

[0002] In recent years, haze has shown a trend of spreading across the country. In most parts of the country, PM2.5 seriously interferes with people's daily life and travel, and is harmful to human health. "Preventing and controlling haze" has always been a matter of concern to the government and the public. At present, the haze information people receive is the haze data released by the National Environmental Monitoring Center. Although this way of publishing haze information solves people's need to understand haze monitoring data, there are problems such as untimely data release, high cost of professional equipment, and rough and inaccurate coverage granularity. It is impossible to "understand haze and prevent haze" in any place in a timely manner.

[0003] The shortcomings of traditional smog monitoring equipment are that it will be affected by the layout of urban buildings, factories, greening, etc., and the wind direction, temperature, terrain, time difference, etc. will cause errors in the data. Therefore, the monitoring results of traditional smog monitoring equipment are different from the smog data of an individual's actual location. In addition, traditional smog monitoring equipment only collects data and cannot process smog data in a timely manner on the device side, and cannot provide smog source information in a timely manner, making it difficult to prevent and control smog.

[0004] Therefore, it is urgent to find an accurate, low-cost and widely applicable method to solve the above problems derived from traditional haze monitoring equipment. In view of the many practical problems caused by traditional haze monitoring equipment, the present invention proposes a haze pollution source tracing system and its source tracing method based on intelligent video monitoring processing. Summary of the invention

[0005] The purpose of the present invention is to provide a haze pollution source investigation system based on video monitoring, which solves the problem that the haze monitoring equipment in the prior art is high in cost and cannot provide haze source information in time.

[0006] Another object of the present invention is to provide a method for tracing the source of the haze pollution source tracing system, which solves the problem in the prior art that the haze monitoring system cannot provide haze source information in a timely manner.

[0007] The technical solution adopted by the present invention is a smog pollution source investigation system based on video monitoring, including a bottom circuit connection substrate, the upper end of the bottom circuit connection substrate is fixedly connected to an image sensor, the upper end of the image sensor is fixedly connected to the lower end of an image sensing channel, a focus motor is fixedly installed on the upper end of the image sensing channel, the upper end of the focus motor is fixedly connected to a lens group, and the lens group is composed of a number of cameras; an infrared filter is fixedly connected between the upper part of the image sensor and the lower part of the lens group; an arc-shaped fixing frame is welded to the upper end of the focus motor, and the lens group is fixedly installed on the arc-shaped fixing frame.

[0008] The present invention is also characterized in that:

[0009] The infrared filter is clamped on the outer wall of the image sensing channel, the bottom of the infrared filter is fixedly connected to the middle circuit connection substrate, and the lower end of the middle circuit connection substrate is fixedly installed on the outer side of the bottom circuit connection substrate.

[0010] The upper part of the bottom circuit connection substrate is a symmetrical oblique rod, and the lower part is a symmetrical L-shape. The upper and lower parts of the bottom circuit connection substrate are integrally formed, and an opening is provided between the symmetrical L-shaped bottom ends; connection ears are fixedly provided on both sides of the bottom circuit connection substrate, and the lower end of the middle circuit connection substrate is fixedly mounted on the connection ears by screws.

[0011] Another technical solution adopted by the present invention is a source tracing method of a haze pollution source tracing system based on video monitoring, firstly, by using several cameras on a lens group to collect street information in different time periods at several locations respectively, the collected video information is cropped to generate pictures of the same time point at different locations; then the picture is obtained by using a digital image method to obtain pixel point data information of the image, and the pixel point information of the picture is gradient processed to obtain haze pollution parameters; the haze pollution parameters of different locations are processed by Poisson's equation to obtain the haze pollution concentration of each location at a certain time point; the haze pollution parameters at different time points at the same location are derived to obtain the change value of the haze pollution parameter; then the value after the second-order derivative of the haze pollution concentration and the change value of the haze pollution parameter are fitted to obtain the shooting position of the picture represented by the maximum value of both values, which is the pollution source of the haze.

[0012] The present invention is also characterized in that:

[0013] The specific steps include:

[0014] Step 1, extracting real-time street video information: using several cameras on the lens group to collect video data, collect videos of several different locations in the same time period t, then cut the collected videos into picture information in the time period t, and use the digital image method to convert each pixel of the image into a data;

[0015] Step 2: Based on the pixel data information obtained in step 1, the information is integrated and the Roberts operator is used to perform gradient processing on the pixel data information of the image to obtain the haze pollution parameter a at the same location in different time periods t. The haze pollution parameter a is derived to obtain the change value A of the haze pollution parameter;

[0016] Step 3: Based on the haze pollution parameter a obtained in step 2, the Poisson equation is used to obtain the haze concentration m at a certain time point in each location;

[0017] Step 4: by taking the second-order derivative of the haze concentration m obtained in step 3 in the time series to obtain the change value B of the haze concentration in a certain period of time, the change value B of the haze concentration in a certain period of time and the change value A of the haze pollution parameter obtained in step 2 are fitted, and it is found that the shooting position of the picture represented by the maximum value of both is the source of the haze in this position area.

[0018] Step 2 specifically includes the following steps:

[0019] Step 2.1, establish a coordinate system with the lower left point of the image as the origin and the two adjacent and perpendicular sides as the x-axis and y-axis respectively;

[0020] Step 2.2: Convert the digital image information obtained in step 1 into an image function f(x,y). The gradient at point (x,y) is a vector with magnitude and direction. Gx and Gy represent the gradients in the x and y directions respectively. Then, according to the definition of the partial equation, as well as The vector expression of the gradient at a point (x, y) is:

[0021]

[0022] The magnitude of this vector is:

[0023]

[0024] The direction angle is:

[0025]

[0026] For digital images, it is equivalent to finding the gradient of a two-dimensional discrete function. A coordinate axis Oxu is established in the image. Let a point I(i,y) in the coordinate system, G(x,y) represents the gradient on the coordinate axis. Then dx(i,j) represents the gradient on the X axis. Similarly, dy(i,j) is the gradient on the Y axis. The gradient at this point is calculated as follows:

[0027] G(x,y)=dx(i,j)+dy(i,y)

[0028] dx(i,j)=I(i+1,j)-I(i,j)

[0029] dy(i,j)=I(i,j+1)-I(i,j) (4)

[0030] In digital images, differences are more often used to approximate derivatives. The gradient approximation expression is as follows:

[0031] G x =f(x,y)-f(x-1,y)

[0032] G y =f(x,y)-f(x,y-1) (5)

[0033] The direction of the gradient is the direction in which the function f(x,y) changes fastest, then G x G is the direction in which the gradient changes fastest on the X axis. y It is the direction in which the gradient changes fastest on the Y axis;

[0034] Through the mathematical one-dimensional integration method Δf=f(x+1)-f(x), for a two-dimensional digital image f(x,y), it is necessary to complete the differentiation in the XY direction. is the partial differential of the image on the X-axis, is the partial differential of the image on the Y axis, the formula is as follows:

[0035]

[0036] Calculate their partial differentials in the X and Y directions respectively, and finally get the gradient

[0037] According to the gradient vector values ​​in the X and Y directions, xGradient is the gradient vector of this point on the X axis, and yGradient is the gradient vector of this point on the Y axis, we can get the following important gradient parameter amplitude magnitude, and the calculation formula is as follows:

[0038]

[0039] From this, we can derive the gradient of the image;

[0040] Step 2.3: Calculate the haze pollution parameter a

[0041] According to step 2.2, the gradient of the image is obtained, and then the value of the haze pollution parameter a1 in the image at a certain time point is obtained by fitting it with the original image and the gradient of the image. Then, this method is used to take images of the same location but in different time periods t for processing to obtain parameters a2, a3, ..., an, and then the derivative is performed to obtain the change value A of the haze pollution parameter.

[0042] The specific contents of step 3 are as follows:

[0043] The formula for calculating haze concentration using Poisson's equation is as follows:

[0044] Δf = -ρ (8)

[0045] Δ represents the Laplace operator. The Laplace transform of the two-dimensional image function is an isotropic second-order derivative, defined as:

[0046]

[0047] In order to be more suitable for digital image processing, the equation is expressed in discrete form:

[0048]

[0049] Use Poisson's equation to get the haze concentration m in the image.

[0050] The beneficial effects of the present invention are as follows: the present invention captures image information within the same time period of the video information obtained by the camera, and then converts the image into pixel data information, and then uses the Poisson equation to obtain a concentration value based on the change in the pixel data information as a parameter, and then the concentration value is derived twice to obtain the haze source information, which is more conducive to relevant departments to find the source of haze, thereby achieving the effect of accurate haze control; the present invention detects the source of haze through image processing methods, and the obtained data is more timely and easy to operate. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 It is a structural schematic diagram of a haze pollution source tracing system based on image processing according to the present invention;

[0052] Figure 2 It is a flow chart of a method for finding the source of haze pollution in a haze pollution source finding system based on image processing according to the present invention;

[0053] Figure 3 The present invention discloses a flow chart of a haze pollution source investigation system for gradient processing of images based on image processing.

[0054] In the figure: 1. Lens group, 2. Focus motor, 3. Infrared filter, 4. Image sensing channel, 5. Image sensor, 6. Bottom line connection substrate, 7. Middle line connection substrate. DETAILED DESCRIPTION

[0055] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0056] A smog pollution source tracing system based on video surveillance, such as Figure 1As shown, it includes a bottom circuit connecting substrate 6, the upper end of the bottom circuit connecting substrate 6 is fixedly connected to the image sensor 5, the upper end of the image sensor 5 is fixedly connected to the lower end of the image sensing channel 4, the upper end of the image sensing channel 4 is fixedly installed with a focus motor 2, the upper end of the focus motor 2 is fixedly connected to a lens group 1, and the lens group 1 is composed of a number of cameras; an infrared filter 3 is fixedly connected between the upper part of the image sensor 5 and the lower part of the lens group 1; an arc-shaped fixing frame is welded to the upper end of the focus motor 2, and the lens group 1 is fixedly installed on the arc-shaped fixing frame.

[0057] like Figure 1 As shown, the infrared filter 3 is clamped on the outer wall of the image sensing channel 4, the bottom of the infrared filter 3 is fixedly connected to the middle circuit connection substrate 7, and the lower end of the middle circuit connection substrate 7 is fixedly installed on the outside of the bottom circuit connection substrate 6. The upper part of the bottom circuit connection substrate 6 is a symmetrical oblique rod, and the lower part is a symmetrical L-shaped. The upper and lower parts of the bottom circuit connection substrate 6 are integrally formed, and an opening is provided between the symmetrical L-shaped bottom ends; connection ears are fixedly provided on both sides of the bottom circuit connection substrate 6, and the lower end of the middle circuit connection substrate 7 is fixedly installed on the connection ears by screws.

[0058] A method for locating the source of haze pollution based on video monitoring. Figure 2 and 3 As shown, firstly, by using several cameras on the lens group 1 to collect street information in different time periods at several locations, the collected video information is cropped to generate pictures of different locations at the same time point; then, the digital image method is used to obtain the pixel point data information of the image, and the pixel point information of the image is gradient processed to obtain the haze pollution parameters; the haze pollution parameters of different locations are processed by Poisson's equation to obtain the haze pollution concentration of each location at a certain time point; the haze pollution parameters of the same location at different time points are derived to obtain the change value of the haze pollution parameter; then, the value after the second-order derivative of the haze pollution concentration and the change value of the haze pollution parameter are fitted to obtain that the shooting position of the picture represented by the maximum value of both values ​​is the pollution source of the haze.

[0059] like Figure 2 and 3 As shown, a method for locating the source of haze pollution based on video monitoring specifically includes the following steps:

[0060] Step 1, extracting real-time street video information: using several cameras on lens group 1 to collect video data, collect videos of several different locations in the same time period t, then cut the collected videos into image information in the time period t, and use digital imaging to convert each pixel of the image into a data;

[0061] Step 2: Based on the pixel data information obtained in step 1, the information is integrated and the Roberts operator is used to perform gradient processing on the pixel data information of the image to obtain the haze pollution parameter a at the same location in different time periods t. The haze pollution parameter a is derived to obtain the change value A of the haze pollution parameter;

[0062] Step 3: Based on the haze pollution parameter a obtained in step 2, the Poisson equation is used to obtain the haze concentration m at a certain time point in each location;

[0063] Step 4: by taking the second-order derivative of the haze concentration m obtained in step 3 in the time series to obtain the change value B of the haze concentration in a certain period of time, the change value B of the haze concentration in a certain period of time and the change value A of the haze pollution parameter obtained in step 2 are fitted, and it is found that the shooting position of the picture represented by the maximum value of both is the source of the haze in this position area.

[0064] Step 2 specifically includes the following steps:

[0065] Step 2.1, establish a coordinate system with the lower left point of the image as the origin and the two adjacent and perpendicular sides as the x-axis and y-axis respectively;

[0066] Step 2.2: Convert the digital image information obtained in step 1 into an image function f(x,y). The gradient at point (x,y) is a vector with magnitude and direction. Gx and Gy represent the gradients in the x and y directions respectively. Then, according to the definition of the partial equation, as well as The vector expression of the gradient at a point (x, y) is:

[0067]

[0068] The magnitude of this vector is:

[0069]

[0070] The direction angle is:

[0071]

[0072] For digital images, it is equivalent to finding the gradient of a two-dimensional discrete function. The coordinate axis Oxy is established in the image. Let a point I(i,y) in the coordinate system, G(x,y) represents the gradient on the coordinate axis, then dx(i,j) represents the gradient on the X axis, and dy(i,j) is the gradient on the Y axis. The gradient at this point is as follows:

[0073] G(x,y)=dx(i,j)+dy(i,y)

[0074] dx(i,j)=I(i+1,j)-I(i,j)

[0075] dy(i,j)=I(i,j+1)-I(i,j) (4)

[0076] In digital images, differences are more often used to approximate derivatives. The gradient approximation expression is as follows:

[0077] G x =f(x,y)-f(x-1,y)

[0078] G y =f(x,y)-f(x,y-1) (5)

[0079] The direction of the gradient is the direction in which the function f(x,y) changes fastest, then G x G is the direction in which the gradient changes fastest on the X axis. y It is the direction where the gradient changes fastest on the Y axis. When there is an edge in the image, there must be a larger gradient value. On the contrary, when there is a relatively smooth part in the image, the gray value changes less, and the corresponding gradient is also smaller.

[0080] Through the mathematical one-dimensional integration method Δf=f(x+1)-f(x), for a two-dimensional digital image f(x,y), it is necessary to complete the differentiation in the XY direction. is the partial differential of the image on the X-axis, is the partial differential of the image on the Y axis, the formula is as follows:

[0081]

[0082] Calculate their partial differentials in the X and Y directions respectively, and finally get the gradient

[0083] Image differentiation (gradient calculation) is an important intermediate step in image edge extraction. According to the gradient vector values ​​in the X and Y directions, xGradient is the gradient vector of this point on the X axis, and yGradient is the gradient vector of this point on the Y axis. The following important gradient parameter amplitude magnitude can be obtained. The calculation formula is as follows:

[0084]

[0085] From this, we can derive the gradient of the image;

[0086] Step 2.3: Calculate the haze pollution parameter a

[0087] According to step 2.2, the gradient of the image is obtained, and then the value of the haze pollution parameter a1 in the image at a certain time point is obtained by fitting it with the original image and the gradient of the image. Then, this method is used to take images of the same location but in different time periods t for processing to obtain parameters a2, a3, ..., an, and then the derivative is performed to obtain the change value A of the haze pollution parameter.

[0088] The specific contents of step 3 are as follows:

[0089] The formula for calculating haze concentration using Poisson's equation is as follows:

[0090] Δf=-ρ (8)

[0091] Δ represents the Laplace operator, which is the simplest isotropic differential operator with rotation invariance. The Laplace transform of the two-dimensional image function is an isotropic second-order derivative, defined as:

[0092]

[0093] In order to be more suitable for digital image processing, the equation is expressed in discrete form:

[0094]

[0095] In addition, if a bright spot appears in a darker area of ​​the image, the Laplace operation will make the bright spot brighter. Because the edges in the image are those areas where the grayscale jumps, the Laplace sharpening template is very useful in edge detection. It is difficult for general enhancement techniques to determine the position of the edge line for steep edges and slowly changing edges. However, this operator can be determined by the zero-crossing point between the positive and negative peaks of the second differential, and is more sensitive to isolated points or endpoints. Therefore, it is particularly suitable for occasions where the purpose is to highlight isolated points, isolated lines or line endpoints in the image. Like the gradient operator, the Laplace operator will also enhance the noise in the image. Sometimes when using the Laplace operator for edge detection, the image can be smoothed first. This can be used to obtain the concentration of the image.

[0096] Use Poisson's equation to get the haze concentration m in the image.

Claims

1. A method for locating the source of haze pollution based on video monitoring, characterized in that: The video monitoring smog pollution source investigation system comprises a bottom circuit connection substrate (6), the upper end of the bottom circuit connection substrate (6) is fixedly connected to an image sensor (5), the upper end of the image sensor (5) is fixedly connected to the lower end of an image sensing channel (4), the upper end of the image sensing channel (4) is fixedly mounted with a focus motor (2), the upper end of the focus motor (2) is fixedly connected to a lens group (1), and the lens group (1) is composed of a plurality of cameras; an infrared filter (3) is fixedly connected between the upper part of the image sensor (5) and the lower part of the lens group (1); an arc-shaped fixing frame is welded to the upper end of the focus motor (2), and the lens group (1) is fixedly mounted on the arc-shaped fixing frame; First, by using a plurality of cameras on the lens group (1) to collect street information at a plurality of locations in different time periods, the collected video information is cut to generate images of the same time point at different locations; then, the image pixel data information is obtained from the image using a digital image method, and the pixel information of the image is gradient processed to obtain haze pollution parameters; the haze pollution parameters of different locations are processed using the Poisson equation to obtain haze pollution concentrations at a certain time point at each location; the haze pollution parameters at different time points at the same location are derived to obtain the change value of the haze pollution parameters; then, the value after the second-order derivative of the haze pollution concentration and the change value of the haze pollution parameter are fitted to obtain that the shooting location of the image represented by the maximum value of both values ​​is the pollution source of the haze; The specific steps include: Step 1, extracting real-time street video information: using a plurality of cameras on the lens group (1) to collect video data, collecting videos of a plurality of different locations within the same time period t, then cutting the collected videos into image information within the time period t, and using a digital image method to convert each pixel of the image into a data; Step 2: Based on the pixel data information obtained in step 1, the information is integrated and the Roberts operator is used to perform gradient processing on the pixel data information of the image to obtain the haze pollution parameter a at the same location in different time periods t. The haze pollution parameter a is derived to obtain the change value A of the haze pollution parameter; Step 3: Based on the haze pollution parameter a obtained in step 2, the Poisson equation is used to obtain the haze concentration m at a certain time point in each location; Step 4: by performing the second-order derivation of the haze concentration m obtained in step 3 in the time series to obtain the change value B of the haze concentration in a certain period of time, the change value B of the haze concentration in a certain period of time and the change value A of the haze pollution parameter obtained in step 2 are fitted, and it is found that the shooting position of the picture represented by the maximum value of both values ​​is the haze source in this position area; The step 2 specifically includes the following steps: Step 2.1, establish a coordinate system with the lower left point of the image as the origin and the two adjacent and perpendicular sides as the x-axis and y-axis respectively; Step 2.2: Convert the digital image information obtained in step 1 into an image function f(x,y). The gradient at point (x,y) is a vector with magnitude and direction. Gx and Gy represent the gradients in the x and y directions respectively. Then, according to the definition of the partial equation, as well as The vector expression of the gradient at a point (x, y) is: The magnitude of this vector is: The direction angle is: For digital images, it is equivalent to finding the gradient of a two-dimensional discrete function. The coordinate axis Oxy is established in the image. Let a point I(i,y) in the coordinate system, G(x,y) represents the gradient on the coordinate axis, then dx(i,j) represents the gradient on the X axis, and dy(i,j) is the gradient on the Y axis. The gradient at this point is as follows: G(x,y)=dx(i,j)+dy(i,y) dx(i,j)=I(i+1,j)-I(i,j) dy(i,j)=I(i,j+1)-I(i,j) (4) In digital images, differences are more often used to approximate derivatives. The gradient approximation expression is as follows: G x =f(x,y)-f(x-1,y) G y =f(x,y)-f(x,y-1) (5) The direction of the gradient is the direction in which the function f(x,y) changes fastest, then G x G is the direction in which the gradient changes fastest on the X axis. y It is the direction in which the gradient changes fastest on the Y axis; Through the mathematical one-dimensional integration method Δf=f(x+1)-f(x), for a two-dimensional digital image f(x,y), it is necessary to complete the differentiation in the XY direction. is the partial differential of the image on the X-axis, is the partial differential of the image on the Y axis, the formula is as follows: Calculate their partial differentials in the X and Y directions respectively, and finally get the gradient According to the gradient vector values ​​in the X and Y directions, xGradient is the gradient vector of this point on the X axis, and yGradient is the gradient vector of this point on the Y axis, we can get the following important gradient parameter amplitude magnitude, and the calculation formula is as follows: From this, we can derive the gradient of the image; Step 2.3: Calculate the haze pollution parameter a According to the step 2.2, the gradient of the image is obtained, and then the value of the haze pollution parameter a1 in the image at a certain time point is obtained by fitting with the original image and the gradient of the image. Then, this method is used to take images of the same location but in different time periods t for processing to obtain parameters a2, a3, ..., an, and then the derivative is performed to obtain the change value A of the haze pollution parameter.

2. The method for tracing the source of haze pollution based on video monitoring according to claim 1 is characterized in that: The specific contents of step 3 are as follows: The formula for calculating haze concentration using Poisson's equation is as follows: Δf=-ρ (8) Δ represents the Laplace operator. The Laplace transform of the two-dimensional image function is an isotropic second-order derivative, defined as: In order to be more suitable for digital image processing, the equation is expressed in discrete form: Use Poisson's equation to get the haze concentration m in the image.

3. The method for tracing the source of haze pollution based on video monitoring according to claim 1 or 2, characterized in that: The infrared filter (3) is clamped on the outer wall of the image sensing channel (4), the bottom of the infrared filter (3) is fixedly connected to the middle circuit connection substrate (7), and the lower end of the middle circuit connection substrate (7) is fixedly mounted on the outside of the bottom circuit connection substrate (6).

4. The method for tracing the source of haze pollution based on video monitoring according to claim 3 is characterized in that: The upper part of the bottom circuit connection substrate (6) is a symmetrical oblique rod, and the lower part is a symmetrical L-shape. The upper part and the lower part of the bottom circuit connection substrate (6) are integrally formed, and an opening is provided between the symmetrical L-shaped bottom ends. Connection ears are fixedly provided on both sides of the bottom circuit connection substrate (6), and the lower end of the middle circuit connection substrate (7) is fixedly mounted on the connection ears by means of screws.

Images