Image defogging automatic processing method based on artificial intelligence

Abstract

The invention belongs to the technical field of image defogging automatic processing, and relates to an image defogging automatic processing method based on artificial intelligence. According to the method, the image data set of each specified power transmission line tower to which each area belongs in each shooting is acquired, and the technologies of polarization analysis, defogging processing, image fusion and the like are further performed on the image data set, so that the defogging image of each specified power transmission line tower to which each area belongs is obtained, the influence of a complex environment on target detection is effectively reduced, and the target detection accuracy is improved. Therefore, the state of the power transmission line channel is clear and visible in various fog periods, the state management and control capability of the important power transmission channel is improved, and a more accurate and objective evaluation result is obtained by analyzing the defogging effect evaluation of each specified power transmission line tower defogging image to which each region belongs and outputting and feeding back the defogging effect evaluation. More accurate and reliable information is provided for subsequent image analysis, processing and decision making, and conditions of towers and lines can be observed more clearly.

Description

Technical Field

[0001] The invention belongs to the technical field of automatic image defogging processing, and relates to an automatic image defogging processing method based on artificial intelligence. Background Art

[0002] In the natural environment, due to the presence of particulate matter, water vapor, smoke, etc. in the atmosphere, the target objects in the image are often obscured by haze, resulting in a decrease in image quality and blurred visual effects. Especially in severe weather conditions, such as haze and rainy days, the visibility and clarity of the image will be seriously affected, which brings difficulties to image processing and recognition. Transmission lines are the infrastructure for realizing power transmission and are the key to ensuring safe and stable power supply. However, transmission lines are often erected outdoors, in a harsh natural environment, and distributed over a wide range. As a result, the line channel images obtained by the visual inspection device of the transmission line are often obscured by fog, the image quality is degraded, and the transmission channel status and the icing status of the conductors cannot be fully seen, which affects the accurate analysis of the line operation status, makes it difficult to find hidden dangers, and is easy to misjudge the operation status. Therefore, the automatic image defogging processing system for transmission lines is of great significance and role.

[0003] In the prior art, there are also some related solutions involving image defogging processing, for example, the invention patent application for an automatic image defogging processing method, system, terminal and storage medium with Chinese patent publication number CN117392009A, which includes: using an image classification model to classify the fog concentration of the image to be processed, if the fog concentration of the image to be processed is low, then using a defogging algorithm to defog the image to be processed, if the fog concentration of the image to be processed is high, then adding an optical lens to perform optical defogging, and color restoration of the image obtained by optical defogging. The present invention uses an image classification model to classify the fog concentration level of the image to be processed, automatically enables a matching defogging processing strategy based on the classification as the image to be processed, realizes automatic defogging of the image, and automatically performs color compensation on the image after performing optical defogging, and finally obtains a clear color image.

[0004] Another invention patent application for a polarization-based image defogging method with a Chinese patent publication number of CN118314045A includes: obtaining a fog image and multiple polarization images corresponding to the fog image, and the polarization angles of the multiple polarization images are different; constructing a network model, including a PEN module and a TDN module connected in sequence, and an SRN module connected to the PEN module and the TDN module respectively; inputting multiple polarization images into the network model, and the PEN module obtains the polarization degree of transmitted light, the polarization degree of atmospheric light, the infinite atmospheric light, the characteristic image, and the polarization degree of the characteristic image according to the multiple polarization images; the TDN module obtains the transmission map according to the polarization degree of transmitted light, the polarization degree of atmospheric light, the characteristic image, and the polarization degree of the characteristic image; the SRN module obtains the defogging image according to the infinite atmospheric light, the characteristic image, and the transmission map. By utilizing the polarization characteristics to realize image defogging, the defogging performance for different degrees of haze scenes and real-world haze scenes is improved.

[0005] Although the above schemes have proposed some solutions for image defogging, they still have certain limitations: the existing schemes automatically enable matching defogging processing strategies according to fog concentration level classification, or use polarization characteristics to achieve image defogging, but the existing schemes ignore the quantitative evaluation of the quality of defogging images, so that it is impossible to evaluate the results more accurately and objectively, and it is impossible to deal with it in time when the defogging effect is not ideal, which is not conducive to improving the availability of images, and further, it is not conducive to a clearer observation of the conditions of poles and lines, and it is impossible to timely discover potential safety hazards and ensure the safe operation of transmission lines. Summary of the invention

[0006] In view of this, in order to solve the problems raised in the above background technology, an automatic image defogging method based on artificial intelligence is proposed.

[0007] The objective of the present invention can be achieved through the following technical solutions: The present invention provides an automatic image defogging processing method based on artificial intelligence, comprising: S1, image data set acquisition: the detected transmission line towers are divided according to regions and recorded as designated transmission line towers belonging to each region, and then the designated transmission line towers belonging to each region are photographed at a set short-term interval according to a preset fixed monitoring interval period, to obtain an image data set of the designated transmission line towers belonging to each region photographed each time in the current time period, and recorded as an image data set of the designated transmission line towers belonging to each region photographed each time, wherein the image data set includes pictures corresponding to each polarization angle.

[0008] S2. Image data set analysis: analyzing polarization parameters of each pixel of each designated transmission line tower image of each area captured each time.

[0009] S3. Image data set analysis: Analyze the fog impact index of each pixel point of each designated transmission line tower image belonging to each area, and then obtain each fog impact sub-area corresponding to each designated transmission line tower image belonging to each area.

[0010] S4. Obtaining defogged images: Analyze the atmospheric light intensity and transmission rate of the images of each designated transmission line tower in each area, and accordingly obtain the defogged images corresponding to each polarization angle of each designated transmission line tower in each area, and fuse them to obtain the defogged images of each designated transmission line tower in each area.

[0011] S5. Image defogging effect analysis: obtaining effect data parameters of the defogging images of each designated transmission line tower in each area, and analyzing the defogging effect evaluation of the defogging images of each designated transmission line tower in each area.

[0012] S6, defogging image management: If the defogging effect of the defogging image of each designated transmission line tower in each area is evaluated as excellent, it will be output and fed back, otherwise, S1 will be executed.

[0013] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. The present invention obtains the image data set of each designated transmission line tower in each area captured each time, and further performs polarization analysis, defogging processing and image fusion techniques thereon, thereby obtaining the defogging image of each designated transmission line tower in each area, which helps to obtain higher image clarity and contrast, while retaining more detailed information of the target, effectively reducing the impact of complex environment on target detection, and thus achieving clear visibility of the status of the transmission line channel in various foggy periods, identifying hidden dangers, improving the ability to control the status of important transmission channels, ensuring safe operation, and reducing the cost and difficulty of manual intervention.

[0014] 2. The present invention obtains effect data parameters of defogging images of designated transmission line towers in each area, analyzes the defogging effect evaluation of the defogging images of designated transmission line towers in each area, outputs and feeds back the defogging effect, and quantitatively evaluates the quality of the defogging images, thereby obtaining more accurate and objective evaluation results, which helps to improve the usability of the images, and provide more accurate and reliable information for subsequent image analysis, processing and decision-making, thereby facilitating clearer observation of the conditions of towers and lines, timely discovery of potential safety hazards, and ensuring the safe operation of transmission lines. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for describing the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0016] Figure 1 It is a schematic diagram for implementing the method steps of the present invention. DETAILED DESCRIPTION

[0017] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0018] See also Figure 1 As shown, the present invention provides an automatic image defogging processing method based on artificial intelligence, and the specific steps are as follows: S1, image data set acquisition: the detected transmission line towers are divided according to regions and recorded as designated transmission line towers belonging to each region, and then the designated transmission line towers belonging to each region are photographed at a set short-term interval according to a preset fixed monitoring interval period, to obtain an image data set of the designated transmission line towers belonging to each region photographed each time in the current time period, and it is recorded as the image data set of the designated transmission line towers belonging to each region photographed each time, wherein the image data set includes each picture corresponding to each polarization angle.

[0019] S2. Image data set analysis: analyzing polarization parameters of each pixel of each designated transmission line tower image of each area captured each time.

[0020] As a preferred feasible embodiment, the polarization parameters of each pixel point of each designated power transmission line tower image of each area captured each time include polarization degree and polarization angle.

[0021] As a preferred feasible embodiment, the specific method for obtaining the polarization degree of each pixel point of the image of each designated transmission line tower belonging to each area shot each time is: extracting each picture corresponding to each polarization angle of each designated transmission line tower belonging to each area shot each time, processing it using an image processing library, and obtaining the light intensity value of each pixel point corresponding to each picture corresponding to each polarization angle of each designated transmission line tower belonging to each area shot each time, and then obtaining the average light intensity value of each pixel point corresponding to each polarization angle of each designated transmission line tower belonging to each area shot each time.

[0022] It should be further explained that the specific method for obtaining the average light intensity value of each pixel corresponding to each polarization angle of each designated transmission line tower in each area of ​​each shooting is as follows: according to the calculation formula Get the average light intensity value G of each pixel corresponding to each polarization angle of each designated transmission line tower in each area of ​​each shooting fabhd , where G′ fabhyd is the average light intensity value of the dth pixel in the yth picture corresponding to the hth polarization angle of the bth designated transmission line tower in the ath area taken for the fth time, f=1,2,...,F, f is the number of each shooting, F is the number of shootings, a=1,2,...,i, a is the number of each area, i is the number of areas, b=1,2,...,B, b is the number of each designated transmission line tower, B is the number of designated transmission line towers, h=1,2,...,H, h is the number of each polarization angle, H is the number of polarization angles, y=1,2,...,Y, y is the number of each picture, Y is the number of pictures, d=1,2,...,D, d is the number of each pixel, D is the number of pixels.

[0023] The average light intensity values ​​of each pixel corresponding to each polarization angle of each designated transmission line tower in each area of ​​each shooting are sorted from large to small according to the different polarization angles, and the maximum average light intensity value and the minimum average light intensity value of each pixel of each designated transmission line tower in each area of ​​each shooting at different polarization angles are obtained, which are recorded as Where f = 1, 2, ..., F, f is the number of each shot, F is the number of shots, a = 1, 2, ..., i, a is the number of each area, i is the number of areas, b = 1, 2, ..., B, b is the number of each designated transmission line tower, B is the number of designated transmission line towers, d = 1, 2, ..., D, d is the number of each pixel, D is the number of pixels, according to the calculation formula Get the polarization degree P of each pixel point of each designated transmission line tower image of each area captured each time fabd .

[0024] As a preferred feasible embodiment, the specific method for obtaining the polarization angle of each pixel point of the image of each designated transmission line tower belonging to each area captured each time is: the average light intensity value of each pixel point corresponding to each polarization angle of each designated transmission line tower belonging to each area captured each time is obtained by screening the average light intensity value of each pixel point of each designated transmission line tower belonging to each area captured each time at the polarization angles of 0 degrees, 45 degrees, 90 degrees and 135 degrees. According to the calculation formula Get the polarization angle θ of each pixel point of each designated transmission line tower image in each area captured each time fabd .

[0025] S3. Image data set analysis: Analyze the fog impact index of each pixel point of each designated transmission line tower image belonging to each area, and then obtain each fog impact sub-area corresponding to each designated transmission line tower image belonging to each area.

[0026] As a preferred feasible embodiment, the specific analysis method of the fog influence degree index of each pixel point of each designated transmission line tower image belonging to each area is: extracting the polarization degree and polarization angle of each pixel point of each designated transmission line tower image belonging to each area taken each time, analyzing the fog influence degree index of each pixel point of each designated transmission line tower image belonging to each area Where P( f+1 ) abd ,θ( f+1 ) abd are the polarization degree and polarization angle of the dth pixel point of the image of the bth designated transmission line tower in the ath area taken for the (f+1)th time, respectively. △P0 and △θ0 are the set allowable polarization degree difference and allowable polarization angle difference, respectively.

[0027] It should be further explained that the set permissible polarization degree difference and permissible polarization angle difference may be 0.01 and 0.01 degrees respectively, which means that the polarization degree and polarization angle of the pixel point may have a deviation of one percent.

[0028] As a preferred feasible embodiment, the specific method for obtaining the sub-regions of fog impact degree corresponding to the designated transmission line tower images of each area is as follows: the area enclosed by adjacent pixel points of the designated transmission line tower images of each area with consistent fog impact degree indicators is recorded as the sub-region of fog impact degree corresponding to the designated transmission line tower images of each area, and then the sub-regions of fog impact degree corresponding to the designated transmission line tower images of each area are obtained.

[0029] It needs to be further explained that after obtaining the sub-areas of fog influence degree corresponding to the images of each designated transmission line tower belonging to each area, different degrees of defogging intensity are performed on the sub-areas of fog influence degree corresponding to the images of each designated transmission line tower belonging to each area based on the degree of influence of the fog concentration.

[0030] S4. Obtaining defogged images: Analyze the atmospheric light intensity and transmission rate of the images of each designated transmission line tower in each area, and accordingly obtain the defogged images corresponding to each polarization angle of each designated transmission line tower in each area, and fuse them to obtain the defogged images of each designated transmission line tower in each area.

[0031] As a preferred feasible embodiment, the specific analysis method of the atmospheric light intensity and transmission rate of the images of the designated transmission line towers in each region is as follows: extract the average light intensity value of each pixel corresponding to each polarization angle of each designated transmission line tower in each region photographed each time, screen out the pixel corresponding to each polarization angle of each designated transmission line tower in each region photographed each time with the largest average light intensity value, record it as the high-brightness pixel corresponding to each polarization angle of each designated transmission line tower in each region photographed each time, and further record the average light intensity value of the high-brightness pixel corresponding to each polarization angle of each designated transmission line tower in each region photographed each time as the atmospheric light intensity A of the image of each designated transmission line tower in each region. ab .

[0032] The polarization degree of each pixel of each designated transmission line tower image of each area captured in each time is extracted, and it is brought into a specific calculation formula to calculate the transmission rate t of each designated transmission line tower image of each area. ab .

[0033] It should be further explained that the specific calculation formula is P=k*t, where P is the polarization degree of the pixel, t is the transmission rate, and k is a constant.

[0034] As a preferred feasible embodiment, the specific method of obtaining the defogging image corresponding to each polarization angle of each designated transmission line tower in each area is: according to the calculation formula Get the defogging light intensity value J of each pixel corresponding to each polarization angle of each designated transmission line tower in each area abhd , where G fabhd The average light intensity value of the dth pixel corresponding to the hth polarization angle of the bth designated transmission line tower in the ath area taken for the fth time, h=1, 2, ..., H, h is the number of each polarization angle, and H is the number of polarization angles.

[0035] The image processing library is used to reconstruct the defogging light intensity value of each pixel corresponding to each polarization angle of each designated transmission line tower in each area, so as to obtain the defogging image corresponding to each polarization angle of each designated transmission line tower in each area.

[0036] It needs to be further explained that the specific operation of using the image processing library to reconstruct the defogging light intensity value of each pixel point corresponding to each polarization angle of each designated transmission line tower belonging to each area includes: using the image processing library to create a new image data structure according to the size of the original image corresponding to each polarization angle of each designated transmission line tower belonging to each area, and constructing a corresponding defogging image in the new image data structure according to the position of each pixel point in the original image corresponding to each polarization angle of each designated transmission line tower belonging to each area, and performing color adjustment.

[0037] It needs to be further explained that the specific operation of fusing the dehazed images corresponding to the polarization angles of the designated transmission line towers in each area to obtain the dehazed images of the designated transmission line towers in each area includes: using a feature point matching algorithm to find the corresponding feature points in the dehazed images corresponding to the polarization angles of the designated transmission line towers in each area, calculating the transformation matrix between the images according to the positions of the feature points, including parameters such as translation, rotation and scaling, and applying the transformation matrix to align all images to a common coordinate system.

[0038] The pixel values ​​of the defogging images corresponding to the polarization angles of the designated transmission line towers in each area are weightedly summed according to certain weights to obtain the fused pixel values, and then the defogging images of the designated transmission line towers in each area are obtained.

[0039] The present invention obtains the image data set of each designated transmission line tower in each area captured each time, and further performs polarization analysis, defogging processing and image fusion techniques on it, so as to obtain the defogging image of each designated transmission line tower in each area, which is helpful to obtain higher image clarity and contrast, while retaining more detailed information of the target, effectively reducing the impact of complex environment on target detection, and thus achieving clear visibility of the status of the transmission line channel in various foggy periods, identifying hidden dangers, improving the status control capability of important transmission channel, ensuring safe operation, and reducing the cost and difficulty of manual intervention.

[0040] S5. Image defogging effect analysis: obtaining effect data parameters of the defogging images of each designated transmission line tower in each area, and analyzing the defogging effect evaluation of the defogging images of each designated transmission line tower in each area.

[0041] As a preferred feasible embodiment, the effect data parameters of the defogging images of each designated power transmission line tower in each area include peak signal-to-noise ratio, structural similarity index and information entropy.

[0042] It should be further explained that the specific method for obtaining the peak signal-to-noise ratio of the defogging image of each designated transmission line tower in each area is: grayscale the defogging image of each designated transmission line tower in each area, and then obtain the defogging grayscale image of each designated transmission line tower in each area and the grayscale value of each pixel corresponding to it. abq , where q = 1, 2, ..., Q, q is the number of pixels corresponding to the grayscale image of the specified transmission line tower, Q is the number of pixels corresponding to the grayscale image of the specified transmission line tower. Similarly, the grayscale value Gray′ of each pixel corresponding to the standard grayscale image of each specified transmission line tower in each area can be obtained. abq .

[0043] According to the analytical formula Get the mean square error (MSE) of the dehazed image of each designated transmission line tower in each area ab .

[0044] According to the analytical formula Get the peak signal-to-noise ratio (PSNR) of the dehazed image of each designated transmission line tower in each area ab Gray max is the maximum possible pixel value of the image. For example, for an eight-bit grayscale image, Gray max =255.

[0045] The specific method for obtaining the structural similarity index of the defogging images of the designated transmission line towers in each region is as follows: according to the grayscale value of each pixel corresponding to the defogging grayscale image of the designated transmission line towers in each region and the grayscale value of each pixel corresponding to the standard grayscale image of the designated transmission line towers in each region, the average brightness of the defogging grayscale image of the designated transmission line towers in each region is calculated. And the average brightness of the standard grayscale image of each designated transmission line tower in each area in and And calculate the standard deviation of the defogging grayscale image of each designated transmission line tower in each area And the standard deviation of the standard grayscale image of each designated transmission line tower in each area in and

[0046]

[0047] According to the analytical formula Get the brightness contrast index l of the dehazed image of each designated transmission line tower in each area ab , where μ0 is a set constant.

[0048] According to the analytical formula Obtain the contrast index c of the defogging image of each designated transmission line tower in each area ab , where μ0′ is a set constant.

[0049] According to the analytical formula Obtain the contrast comparison index s of the dehazed image of each designated transmission line tower in each area ab , where μ0″ is a set constant.

[0050] Analyze the structural similarity index SSIM of the dehazed images of each designated transmission line tower in each area ab = l ab *c ab *s ab .

[0051] The specific method for obtaining the information entropy of the defogging image of each designated transmission line tower in each area is as follows: respectively extracting the grayscale value of each pixel point corresponding to the defogging grayscale image of each designated transmission line tower in each area, and further performing statistics on the grayscale value of each pixel point in the defogging grayscale image of each designated transmission line tower in each area to obtain the number m of each grayscale value of the pixel point abz , where z = 1, 2, ..., W, z is the number of each gray value, W is the number of gray values, according to the analysis formula Get the information entropy of the defogging image of each designated transmission line tower in each area ab .

[0052] As a preferred feasible embodiment, the specific analysis method for evaluating the defogging effect of the defogging images of the designated transmission line towers in each area is as follows: extracting the peak signal-to-noise ratio, structural similarity index and information entropy of the defogging images of the designated transmission line towers in each area, which are respectively recorded as PSNR ab 、SSIM ab 、Entropy ab , analyze the defogging effect evaluation coefficient of the defogging images of each designated transmission line tower in each area Among them, PSNR0, SSIM0, and Entropy0 are the peak signal-to-noise ratio threshold, structural similarity index threshold, and information entropy threshold of the dehazed image of the transmission line tower extracted from the information database, respectively.

[0053] The defogging effect evaluation coefficients of the defogging images of each designated transmission line tower in each area are compared with the set defogging effect evaluation coefficient thresholds. If the defogging effect evaluation coefficient of the defogging image of a designated transmission line tower in a certain area is greater than the defogging effect evaluation coefficient threshold, the defogging effect evaluation of the defogging image of the designated transmission line tower in the area is recorded as excellent.

[0054] S6, defogging image management: If the defogging effect of the defogging image of each designated transmission line tower in each area is evaluated as excellent, it will be output and fed back, otherwise, S1 will be executed.

[0055] As a preferred feasible embodiment, the specific operation of outputting and feeding back the defogging images of each designated transmission line tower in each area whose defogging effect is evaluated as excellent is: uploading the defogging images of each designated transmission line tower in each area to the external network platform through the public network 4G channel; otherwise, re-execute S1 until the defogging effect is evaluated as excellent.

[0056] The present invention obtains effect data parameters of defogging images of designated transmission line towers in each area, analyzes the defogging effect evaluation of the defogging images of designated transmission line towers in each area, outputs and feeds back the defogging effect, and quantitatively evaluates the quality of the defogging images, thereby obtaining more accurate and objective evaluation results, which helps to improve the usability of the images, and provide more accurate and reliable information for subsequent image analysis, processing and decision-making, thereby facilitating clearer observation of the conditions of towers and lines, timely discovery of potential safety hazards, and ensuring the safe operation of transmission lines.

[0057] The above contents are merely examples and explanations of the concept of the present invention. Those skilled in the art may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the concept of the invention or exceed the scope defined by the present invention, they shall all fall within the protection scope of the present invention.

Claims

1. An automatic image defogging method based on artificial intelligence, characterized in that: include: S1. Acquisition of image data set: Divide each detected transmission line tower into designated transmission line towers belonging to each region according to the region, and then photograph each designated transmission line tower belonging to each region at a set short-term interval according to a preset fixed monitoring interval period, and obtain an image data set of each designated transmission line tower belonging to each region photographed each time in the current time period, and record it as an image data set of each designated transmission line tower belonging to each region photographed each time, wherein the image data set includes each picture corresponding to each polarization angle; S2, image data set analysis: analyzing the polarization parameters of each pixel of each designated transmission line tower image of each area captured each time; S3, image data set analysis: analyzing the fog impact index of each pixel point of each designated transmission line tower image belonging to each area, and then obtaining each fog impact sub-area corresponding to each designated transmission line tower image belonging to each area; S4, obtaining defogging images: analyzing the atmospheric light intensity and transmission rate of the images of the designated transmission line towers in each region, and obtaining defogging images corresponding to the polarization angles of the designated transmission line towers in each region, and fusing them to obtain defogging images of the designated transmission line towers in each region; S5, image defogging effect analysis: obtaining effect data parameters of the defogging images of each designated transmission line tower in each area, and analyzing the defogging effect evaluation of the defogging images of each designated transmission line tower in each area; S6, defogging image management: If the defogging effect of the defogging image of each designated transmission line tower in each area is evaluated as excellent, it will be output and fed back, otherwise, S1 will be executed.

2. The method for automatic image defogging based on artificial intelligence according to claim 1, characterized in that: The polarization parameters of each pixel point of each designated transmission line tower image of each area captured each time include polarization degree and polarization angle; The effect data parameters of the defogging images of each designated transmission line tower in each area include peak signal-to-noise ratio, structural similarity index and information entropy.

3. The method for automatic image defogging based on artificial intelligence according to claim 2 is characterized in that: The specific method for obtaining the polarization degree of each pixel point of each designated transmission line tower image of each area captured each time is as follows: Extract the pictures corresponding to the polarization angles of the designated transmission line towers in each area of ​​each shooting, process them using the image processing library, obtain the light intensity value of each pixel point corresponding to the pictures corresponding to the polarization angles of the designated transmission line towers in each area of ​​each shooting, and then obtain the average light intensity value of each pixel point corresponding to the polarization angles of the designated transmission line towers in each area of ​​each shooting; The average light intensity values ​​of each pixel corresponding to each polarization angle of each designated transmission line tower in each area of ​​each shooting are sorted from large to small according to the different polarization angles, and the maximum average light intensity value and the minimum average light intensity value of each pixel of each designated transmission line tower in each area of ​​each shooting at different polarization angles are obtained, which are recorded as Where f = 1, 2, ..., F, f is the number of each shot, F is the number of shots, a = 1, 2, ..., i, a is the number of each area, i is the number of areas, b = 1, 2, ..., B, b is the number of each designated transmission line tower, B is the number of designated transmission line towers, d = 1, 2, ..., D, d is the number of each pixel, D is the number of pixels, according to the calculation formula Get the polarization degree P of each pixel point of each designated transmission line tower image of each area captured each time fabd .

4. The method for automatic image defogging based on artificial intelligence according to claim 3 is characterized in that: The specific method for obtaining the polarization angle of each pixel point of each designated transmission line tower image of each area captured each time is as follows: The average light intensity values ​​of each pixel point of each designated transmission line tower in each area of ​​each shooting at each polarization angle corresponding to each pixel point are screened to obtain the average light intensity values ​​of each pixel point of each designated transmission line tower in each area of ​​each shooting at 0 degree, 45 degree, 90 degree and 135 degree polarization angles. According to the calculation formula Get the polarization angle θ of each pixel point of each designated transmission line tower image in each area captured each time fabd .

5. The method for automatic image defogging based on artificial intelligence according to claim 4 is characterized in that: The specific analysis method of the fog influence index of each pixel point of each designated transmission line tower image in each area is as follows: Extract the polarization degree and polarization angle of each pixel point of each designated transmission line tower image in each area taken each time, and analyze the fog influence index of each pixel point of each designated transmission line tower image in each area Where P( f+1 ) abd ,θ( f+1 ) abd are the polarization degree and polarization angle of the dth pixel point of the image of the bth designated transmission line tower in the ath area taken for the (f+1)th time, respectively. △P0 and △θ0 are the set allowable polarization degree difference and allowable polarization angle difference, respectively.

6. The method for automatic image defogging based on artificial intelligence according to claim 5, characterized in that: The specific method for obtaining each sub-region of fog influence degree corresponding to each designated transmission line tower image in each region is as follows: The area surrounded by adjacent pixel points of each designated transmission line tower image belonging to each area with consistent fog impact degree index is recorded as the fog impact degree sub-area corresponding to each designated transmission line tower image belonging to each area, and then the fog impact degree sub-area corresponding to each designated transmission line tower image belonging to each area is obtained.

7. The method for automatic image defogging based on artificial intelligence according to claim 4, characterized in that: The specific analysis method of the atmospheric light intensity and transmission rate of each designated transmission line tower image in each area is as follows: The average light intensity value of each pixel corresponding to each polarization angle of each designated transmission line tower in each area of ​​each shooting is extracted, and the pixel corresponding to each polarization angle of each designated transmission line tower in each area of ​​each shooting with the largest average light intensity value is screened out, and it is recorded as the high-brightness pixel corresponding to each polarization angle of each designated transmission line tower in each area of ​​each shooting, and further the average light intensity value of the high-brightness pixel corresponding to each polarization angle of each designated transmission line tower in each area of ​​each shooting is recorded as the atmospheric light intensity A of the image of each designated transmission line tower in each area. ab ; The polarization degree of each pixel of each designated transmission line tower image of each area captured in each time is extracted, and it is brought into a specific calculation formula to calculate the transmission rate t of each designated transmission line tower image of each area. ab .

8. The method for automatic image defogging based on artificial intelligence according to claim 7, characterized in that: The specific method for obtaining the defogging image corresponding to each polarization angle of each designated transmission line tower in each area is as follows: According to the calculation formula Get the defogging light intensity value J of each pixel corresponding to each polarization angle of each designated transmission line tower in each area abhd , where G fabhd is the average light intensity value of the dth pixel corresponding to the hth polarization angle of the bth designated transmission line tower in the ath area taken the fth time, h=1,2,...,H, h is the number of each polarization angle, and H is the number of polarization angles; The image processing library is used to reconstruct the defogging light intensity value of each pixel corresponding to each polarization angle of each designated transmission line tower in each area, so as to obtain the defogging image corresponding to each polarization angle of each designated transmission line tower in each area.

9. The method for automatic image defogging based on artificial intelligence according to claim 3, characterized in that: The specific analysis method for evaluating the defogging effect of the defogging images of the designated transmission line towers in each area is as follows: The peak signal-to-noise ratio, structural similarity index and information entropy of the dehazed images of each designated transmission line tower in each area are extracted and recorded as PSNR respectively. ab 、SSIM ab 、Entropy ab , analyze the defogging effect evaluation coefficient of the defogging images of each designated transmission line tower in each area Among them, PSNR0, SSIM0, and Entropy0 are the peak signal-to-noise ratio threshold, structural similarity index threshold, and information entropy threshold of the dehazed image of the transmission line tower extracted from the information database, respectively; The defogging effect evaluation coefficients of the defogging images of each designated transmission line tower in each area are compared with the set defogging effect evaluation coefficient thresholds. If the defogging effect evaluation coefficient of the defogging image of a designated transmission line tower in a certain area is greater than the defogging effect evaluation coefficient threshold, the defogging effect evaluation of the defogging image of the designated transmission line tower in the area is recorded as excellent.

10. The method for automatic image defogging based on artificial intelligence according to claim 9, characterized in that: The specific operation of outputting and feeding back the defogging images of each designated transmission line tower in each area evaluated as having excellent defogging effect is as follows: The defogging images of the designated transmission line towers in each area are uploaded to the external network platform through the public network 4G channel. Otherwise, S1 is re-executed until the defogging effect is evaluated as excellent.

Images