Monitoring image real-time defogging system and method based on airfield pavement FOD monitoring system

Through the dark channel prior defog removal algorithm and multi-scale fusion technology, the problem of airport monitoring image defog removal under the influence of haze is solved, and efficient and real-time defog removal effect is achieved, ensuring the stability of the monitoring system in haze weather and accuracy.

CN119027352APending Publication Date: 2024-11-26SHANDONG EAGLE INFORMATION ENG CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411199337.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-29
Publication Date
2024-11-26

AI Technical Summary

Technical Problem

The smog weather causes the image clarity of airport runway monitoring. The existing technology, such as the deep learning defog removal algorithm, is poor in real time and has high computing power requirements, making it difficult to effectively apply in airport monitoring systems.

Method used

The dark channel prior defog removal algorithm is used to calculate the image dark channel diagram through filtering windows of different scales, transmittance estimation and weighted fusion are performed, and guided filtering and atmospheric scattering model inversion are combined to realize real-time defog removal of the monitoring image.

Benefits of technology

Improves the fineness and nature of the defog removal image, while improving computing efficiency and real-time performance, ensuring accurate and reliable monitoring information under haze conditions.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119027352A_ABST
    Figure CN119027352A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of image processing, in particular to a monitoring image real-time defogging system and method based on an airport pavement FOD monitoring system. Comprising the following steps: reading an airport runway monitoring video stream; calculating an image dark channel map by using different filtering windows; calculating coarse transmissivity obtained by different filtering windows; carrying out weighted fusion on the obtained different coarse transmittances; refining the transmissivity after weighted fusion by using guide filtering; and an atmospheric scattering model is used for inversion to obtain a defogged image. According to the invention, an improved dark channel prior defogging algorithm is utilized, filtering windows with different scales are fused to obtain a transmissivity graph, so that the transmissivity graph not only contains the overall characteristics of the image, but also loses detail information of the image as little as possible; according to the method, the calculation efficiency and the real-time performance are remarkably improved while the defogging effect is improved, the stable and reliable defogging effect can be kept under various haze conditions, and it is ensured that the airport runway monitoring system provides more accurate and reliable real-time information in haze weather.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and in particular to a real-time defogging system and method for monitoring images based on an airport pavement FOD monitoring system. Background Art

[0002] The FOD (Foreign Object Debris) monitoring system is a technical system used to detect and identify foreign matter that may damage aircraft. These foreign objects include metal parts, waterproof plastic sheets, gravel, newspapers, bottles, luggage tags, nuts, screws, washers, nails, fuses, machine tools, personal belongings, pens, pencils, buttons, rubber fragments, plastic products, concrete asphalt fragments, ice chips, paper products, animals and plants, etc. The harm caused by FOD is very serious. Foreign objects on the airport pavement can easily be sucked into the engine, causing engine failure. Fragments may also accumulate in mechanical devices, affecting the normal operation of landing gear, wings and other equipment, which will not only damage the aircraft, but also cause casualties, accompanied by huge economic losses. The existing FOD detection system mainly uses radar detection technology and video image recognition technology. The working process of the FOD detection system is as follows:

[0003] The radar continuously scans the runway and triggers an alarm when FOD is detected;

[0004] The monitoring room highlights the location of the discovered target;

[0005] The monitoring room operates the optical equipment pan / tilt to aim it at the target and uses visual confirmation of the FOD scene image;

[0006] Get precise positioning of FOD targets with the help of GPS;

[0007] Determine the harmfulness of FOD targets, inform the removal vehicles, and remove FOD quickly;

[0008] The "toolbox" in the monitoring room automatically records information such as time, climate, FOD characteristics, and FOD location.

[0009] In China, the airport runway foreign object (FOD) detection system independently developed and produced by the Second Research Institute of the Civil Aviation Administration of China has successfully won the bid for the Beijing Daxing International Airport FOD detection system project. The system can automatically complete real-time monitoring, information extraction and automatic evidence collection of foreign objects on the runway. Compared with foreign products, my country's independently developed products are not inferior in technology, far ahead in service, and have very competitive prices. They can also be improved to meet the diverse needs of domestic airports. It optimizes the sensors in the edge-light runway foreign object optoelectronic hybrid detection system, greatly improves the integration, reduces the difficulty of construction, makes later maintenance more convenient, and improves the maintainability of the system. The application of the FOD monitoring system can improve the accuracy and efficiency of airport runway foreign object inspections, shorten the runway occupancy time, and ensure flight safety. Different FOD monitoring systems may differ in performance, characteristics and applicable scenarios. Airports can choose a suitable system according to their own needs and actual conditions.

[0010] The safe operation of airport runways is one of the key elements of air transportation. However, haze weather often seriously affects runway visibility, increasing the risk of flight delays and aviation accidents. Haze is mainly composed of water vapor and suspended particles in the air, which can significantly reduce the clarity of images captured by cameras, making it difficult for ground monitoring systems to provide accurate and reliable real-time information.

[0011] Image dehazing methods rely on computer vision and image processing techniques to restore the clarity of captured images. Such methods include traditional image processing techniques such as histogram equalization, contrast enhancement, and image denoising, as well as the dehazing algorithms based on deep learning that have emerged in recent years.

[0012] In recent years, with the rapid development of deep learning and computer vision technology, image defogging technology based on deep learning has received widespread attention. This type of technology can learn the characteristics of haze and defogging rules from a large number of haze images by training deep learning models such as convolutional neural networks, thereby making significant progress in image restoration and defogging effects. However, the real-time performance of image defogging using deep learning is poor and the computing power requirements are very high.

[0013] To address the above issues, we proposed a real-time defogging system and method for airport runway surveillance videos. We can use the dark channel prior defogging algorithm and make improvements on this basis. We use multi-scale fusion technology to perform transmittance estimation and image restoration at different scales to obtain a more refined and natural defogging image. Summary of the invention

[0014] The object of the present invention is to provide a real-time defogging system and method for monitoring images based on an airport pavement FOD monitoring system, so as to solve the problems raised in the above-mentioned background technology.

[0015] In order to solve the above technical problems, one of the purposes of the present invention is to provide a real-time defogging method for monitoring images based on an airport pavement FOD monitoring system, comprising the following steps:

[0016] S1, read the airport runway monitoring video stream;

[0017] S2, using different filter windows to calculate the dark channel map of the image;

[0018] S3, calculating the rough transmittance obtained by different filter windows;

[0019] S4, performing weighted fusion on the different rough transmittances obtained;

[0020] S5, using guided filtering to refine the transmittance after weighted fusion;

[0021] S6. The dehazed image is obtained by inverting the atmospheric scattering model.

[0022] As a further improvement of the technical solution, in step S1, the specific steps of reading the airport runway monitoring video stream are as follows:

[0023] S1.1. Establish connection: obtain the storage location or network address of the monitoring video stream in the airport pavement FOD monitoring system; use the corresponding interface or library to establish a connection with the video stream data source;

[0024] S1.2. Initialize the video stream reader: create a video stream reader object according to the selected video processing library or framework; set the relevant parameters of the video stream;

[0025] S1.3, start reading video frames: enter the loop, read the video stream frame by frame; use the method of the video stream reader to obtain the image data of the current frame; perform preliminary processing on the read image data;

[0026] S1.4, frame data processing: perform necessary preprocessing on each frame of image; the frame data can be stored in a buffer or data structure for subsequent analysis and processing;

[0027] S1.5, Synchronization and exception handling: handle the synchronization of video streams, ensure the accuracy of frame order and timestamps; monitor possible anomalies during the reading process and take appropriate error handling measures;

[0028] S1.6, continue reading until the end: continue to loop and read the video frames until the end mark of the video stream is reached or the condition for stopping reading is met;

[0029] S1.7. Resource release: After completing the reading of the video stream, release the resources related to the video stream in time to avoid resource leakage.

[0030] As a further improvement of the technical solution, in step S2, the specific steps of using different filter windows to calculate the dark channel map of the image are as follows:

[0031] S2.1, read the video stream, and process the images in the video stream frame by frame;

[0032] S2.2, traverse the single frame image and convert the image into grayscale, and the grayscale value of each pixel is equal to the minimum value of the three-channel color values ​​of each pixel R, G, and B;

[0033] S2.3. Select filter windows of f1*f1, f2*f2, ..., fn*fn respectively to perform minimum value filtering on the grayscale window obtained in the previous step to obtain the dark channel image of the image corresponding to the filter window.

[0034] As a further improvement of the technical solution, in step S2.3, the specific steps of performing minimum value filtering on the grayscale window are as follows:

[0035] S2.3.1. Determine the size of the filter window, for example, fn*fn;

[0036] S2.3.2. For each pixel in the image, determine the range of its neighborhood with the pixel as the center;

[0037] S2.3.3. Find the minimum value among all pixel values ​​in the neighborhood;

[0038] S2.3.4, assigning the minimum value to the central pixel as the pixel value after filtering;

[0039] S2.3.5. Repeat steps S2.3.2 to S2.3.4 to process each pixel in the image until the entire image is filtered.

[0040] As a further improvement of the technical solution, in step S3, the specific steps of calculating the rough transmittance obtained by different filter windows are as follows:

[0041] S3.1, traverse the dark channel images of n filter windows to obtain the positions of the pixels with the largest pixel values ​​in the top 5% of the entire image;

[0042] S3.2, find the pixel points at the corresponding positions in the first step on the foggy image, and calculate the average value of the R, G, and B channel color values ​​of these pixel points as the atmospheric light value A;

[0043] S3.3, using the following formula to calculate the dark channel image corresponding to the n filter windows obtained in step S2 to obtain a rough transmittance; the calculation formula for the rough transmittance is:

[0044]

[0045] in, is the rough transmittance of the dark channel image, w is the transmittance adjustment parameter, A represents the atmospheric light value, c refers to different color channels, I c (y) represents the color value of color channel c of the foggy image at position y, A c represents the color value of atmospheric light in color channel c, and Ω(x) represents the neighborhood centered on pixel x.

[0046] As a further improvement of the technical solution, in step S4, the specific steps of weighted fusion of the obtained different rough transmittances are as follows:

[0047] The rough transmittance obtained in step S3 is weighted fused using the following formula:

[0048] t(x)=d1t f1 (x)+d2t f2 (x)+...+d n t fn (x)

[0049] Among them, t(x) is the transmittance after weighted fusion, t f1 (x) represents the transmittance obtained using the f1*f1 window, t f2 (x) represents the transmittance obtained using the f2*f2 window, t fn (x) represents the transmittance obtained using the fn*fn window; d1, d2...d n t f1 (x), t f2 (x)...t fn (x) corresponding weight, and d1+d2+...+d n =1.

[0050] As a further improvement of the technical solution, in step S5, the specific steps of using guided filtering to refine the transmittance after weighted fusion are as follows:

[0051] S5.1. The transmittance map is used as the input image, and the grayscale image of the foggy image is used as the guide image of the guided filtering. The refined transmittance map is obtained by guided filtering, where the formula is as follows:

[0052]

[0053] Where O represents the output image, G represents the guide image, i represents the index of the pixel, and w k represents the local window centered at pixel k, a k and b kRepresents the coefficient of the linear function when the local window is located at k;

[0054] Among them, a k and b k The solution is as follows:

[0055]

[0056] in, Represents images G and I in local window w k The covariance within Represents the image G in the local window w k The variance within Represents image I in local window w k The mean value in Represents the mean of image G in the local window.

[0057] As a further improvement of the present technical solution, in the step S5.1, the transmittance map of the input image comes from the step S4, and the final transmittance map is obtained by weighted fusion of the different rough transmittances processed in the step S3; the grayscale map of the foggy image as the guide image comes from the step S2.2, in which the single frame image is traversed and grayscaled, and the grayscale map of the foggy image is selected.

[0058] As a further improvement of the technical solution, in step S6, the specific steps of obtaining the defogging image by inverting the atmospheric scattering model are as follows:

[0059] S6.1. Using the atmospheric scattering model formula, the fog-free image is inverted. The formula is as follows:

[0060] I(x)=J(x)*t(x)+A[1-t(x)]

[0061] Among them, I(x) represents the foggy image, J(x) refers to the fog-free image, t(x) refers to the transmittance after weighted fusion, and A refers to the atmospheric light value, that is, the global atmospheric light intensity; the inversion formula is:

[0062]

[0063] The atmospheric light value A has been obtained in step S3.2, and the transmittance t(x) after weighted fusion is obtained in step S4, so the fog-free image can be obtained;

[0064] S6.2. Display the fog-free video stream on the system interface.

[0065] The second object of the present invention is to provide a real-time defogging system for monitoring images based on an airport pavement FOD monitoring system, which is used to implement the steps of the real-time defogging method for monitoring images based on an airport pavement FOD monitoring system, including:

[0066] Video stream reading module, used to read the airport runway monitoring video stream from the airport pavement FOD monitoring system;

[0067] A dark channel map calculation module is used to calculate the dark channel map of the image using different filter windows;

[0068] A rough transmittance calculation module is used to calculate the rough transmittance obtained by different filter windows;

[0069] Transmittance weighted fusion module, used for weighted fusion of different rough transmittances;

[0070] A transmittance refinement module, used to refine the transmittance after weighted fusion using guided filtering;

[0071] The model inversion module is used to obtain the dehazed image through the inversion of the atmospheric scattering model.

[0072] The third object of the present invention is to provide a real-time defogging operation platform device for monitoring images, including a processor, a memory, and a computer program stored in the memory and running on the processor, and the processor is used to implement the steps of the above-mentioned real-time defogging system and method for monitoring images based on the airport pavement FOD monitoring system when executing the computer program.

[0073] The fourth object of the present invention is to provide a computer-readable storage medium, which stores a computer program, and when the computer program is executed by a processor, it implements the steps of the above-mentioned real-time defogging system and method for monitoring images based on the airport pavement FOD monitoring system.

[0074] Compared with the prior art, the present invention has the following beneficial effects:

[0075] 1. In the real-time defogging system and method for monitoring images based on the airport pavement FOD monitoring system, different filter windows are first used to calculate the dark channel map of the image, and then the rough transmittance obtained by different filter windows is calculated, and the different rough transmittances are weighted fused; then the transmittance is further refined by guided filtering, and finally the defogging image is obtained by inversion using the atmospheric scattering model, and the defogging monitoring video stream is displayed on the system interface;

[0076] 2. In the real-time defogging system and method for monitoring images based on the airport pavement FOD monitoring system, an improved dark channel prior defogging algorithm is used to perform transmittance estimation and image restoration at different scales through multi-scale fusion technology to obtain a more refined and natural defogging image (i.e., the transmittance map is obtained by fusing filter windows of different scales), so that it can contain the overall features of the image and lose as little detail information of the image as possible;

[0077] 3. The real-time defogging system and method for monitoring images based on the airport pavement FOD monitoring system not only improves the defogging effect, but also significantly improves the computing efficiency and real-time performance. It can maintain a stable and reliable defogging effect under various haze conditions, ensuring that the airport runway monitoring system provides more accurate and reliable real-time information in haze weather. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] Figure 1 is an exemplary overall method flow chart of the present invention;

[0079] Figure 2 is an exemplary system architecture schematic diagram of the present invention;

[0080] Figure 3 It is a structural diagram of an exemplary electronic computer platform device in the present invention. DETAILED DESCRIPTION

[0081] 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.

[0082] Example 1

[0083] like Figure 1 As shown, this embodiment provides a real-time defogging method for monitoring images based on an airport pavement FOD monitoring system, comprising the following steps:

[0084] S1. Read the airport runway monitoring video stream (for example, you can use the VideoCapture class of the OpenCV library to read the monitoring video stream); the specific steps are as follows:

[0085] S1.1. Establish connection: obtain the storage location or network address of the monitoring video stream in the airport pavement FOD monitoring system; use the corresponding interface or library to establish a connection with the video stream data source; this may involve operations such as configuring network parameters and verifying access rights;

[0086] S1.2. Initialize the video stream reader: Create a video stream reader object according to the selected video processing library or framework; set relevant parameters of the video stream, such as frame rate, resolution, decoding format, etc., to ensure that the video stream can be read and parsed correctly;

[0087] S1.3, start reading video frames: enter a loop, read the video stream frame by frame; use the video stream reader method to obtain the image data of the current frame; perform preliminary processing on the read image data, such as decoding it into a processable image format (such as RGB, grayscale, etc.);

[0088] S1.4, frame data processing: perform necessary preprocessing on each frame of image, such as cropping, scaling, grayscale, etc., to meet the needs of subsequent processing; the frame data can be stored in a buffer or data structure for subsequent analysis and processing;

[0089] S1.5, Synchronization and exception handling: handle the synchronization issues of video streams to ensure the accuracy of frame order and timestamps; monitor abnormal situations that may occur during the reading process, such as connection interruption, video format mismatch, decoding errors, etc., and take corresponding error handling measures, such as reconnecting, skipping abnormal frames or recording error information;

[0090] S1.6, Continue reading until the end: Continue to loop and read video frames until the end mark of the video stream is reached or the conditions for stopping reading are met; according to actual needs, the time range, frame number limit or other termination conditions for reading can be set;

[0091] S1.7. Resource release: After completing the reading of the video stream, release the resources related to the video stream in time, such as closing the connection, clearing the buffer, etc., to avoid resource leakage.

[0092] S2. Calculate the dark channel map of the image using different filter windows; the specific steps are as follows:

[0093] S2.1, read the video stream, and process the images in the video stream frame by frame;

[0094] S2.2, traverse the single frame image and convert the image into grayscale, and the grayscale value of each pixel is equal to the minimum value of the three-channel color values ​​of each pixel R, G, and B;

[0095] S2.3. Select 3*3, 9*9, and 15*15 filter windows respectively to perform minimum value filtering on the grayscale window obtained in the previous step to obtain the dark channel image of the image corresponding to the filter window.

[0096] Among them, minimum filtering is a filtering algorithm in image processing, which is used to reduce noise in an image or smooth the image. Its basic idea is to take the minimum value of the pixel values ​​in its neighborhood as the new value of the pixel at each pixel position in the image.

[0097] Specifically, in step S2.3, the specific steps of performing minimum value filtering on the grayscale window are as follows:

[0098] S2.3.1. Determine the size of the filter window, such as 3*3, 5*5, etc.;

[0099] S2.3.2. For each pixel in the image, determine the range of its neighborhood with the pixel as the center;

[0100] S2.3.3. Find the minimum value among all pixel values ​​in the neighborhood;

[0101] S2.3.4, assigning the minimum value to the central pixel as the pixel value after filtering;

[0102] S2.3.5. Repeat steps S2.3.2 to S2.3.4 to process each pixel in the image until the entire image is filtered.

[0103] For example, for a 3*3 filter window, assuming that the neighboring pixel values ​​of the current pixel are:

[0104]

[0105] Then the minimum filtering result of the pixel is 12.

[0106] In practical applications, different neighborhood shapes and sizes can be used to meet different needs, and multiple filtering can be performed according to specific circumstances to achieve better results. In addition, there are many optimized algorithms and implementation methods that can improve the efficiency of minimum filtering.

[0107] S3, calculating the rough transmittance obtained by different filter windows; the specific steps are as follows:

[0108] S3.1, traverse the dark channel images of the three filter windows to obtain the positions of the pixels with the largest pixel values ​​in the top 5% of the entire image;

[0109] S3.2, find the pixel points at the corresponding positions in the first step on the foggy image, and calculate the average value of the R, G, and B channel color values ​​of these pixel points as the atmospheric light value A;

[0110] S3.3, using the following formula to calculate the dark channel images corresponding to the three filter windows obtained in step S2 to obtain a rough transmittance; the calculation formula for the rough transmittance is:

[0111]

[0112] in, is the rough transmittance of the dark channel image, w is the transmittance adjustment parameter, where w = 0.95 is the preferred value, A represents the atmospheric light value, c refers to different color channels, and I c (y) represents the color value of color channel c of the foggy image at position y, A c represents the color value of atmospheric light in color channel c, and Ω(x) represents the neighborhood centered on pixel x.

[0113] S4, weighted fusion of different rough transmittances obtained; the specific steps are as follows:

[0114] The rough transmittance obtained in step S3 is weighted fused using the following formula:

[0115] t(x)=d1t f1 (x)+d2t f2 (x)+...+d n t fn (x)

[0116] Among them, t(x) is the transmittance after weighted fusion, t f1 (x) represents the transmittance obtained using the f1*f1 window, t f2 (x) represents the transmittance obtained using the f2*f2 window, t fn (x) represents the transmittance obtained using the fn*fn window; d1, d2...d n t f1 (x), t f2 (x)...t fn (x) corresponding weight, and d1+d2+...+d n =1; combined with the rough transmittance of the dark channel image corresponding to the above three filter windows, we have:

[0117] t(x)=0.3t3(x)+0.4t9(x)+0.3t 15 (x)

[0118] Among them, t3(x) represents the transmittance obtained using a 3*3 window, t9(x) represents the transmittance obtained using a 9*9 window, and t 15 (x) represents the transmittance obtained using a 15*15 window.

[0119] S5. Use guided filtering to refine the transmittance after weighted fusion. The specific steps are as follows:

[0120] S5.1, the transmittance map (from step S4, the final transmittance map is obtained by weighted fusion of different rough transmittances processed in step S3) is used as the input image, and the grayscale image of the foggy image (from step S2.2, in which the single frame image is traversed and grayscaled, and the grayscale image of the foggy image is selected) is used as the guide image of the guided filtering, and the refined transmittance map is obtained by guided filtering, wherein the formula is as follows:

[0121]

[0122] Where O represents the output image, G represents the guide image, i represents the index of the pixel, and w k represents the local window centered at pixel k, a k and b k Represents the coefficient of the linear function when the local window is located at k;

[0123] Among them, a k and b k The solution is as follows:

[0124]

[0125] in, Represents images G and I in local window w k The covariance within Represents the image G in the local window w k The variance within Represents image I in local window w k The mean value in Represents the mean of image G in the local window.

[0126] S6. Obtain the defogging image by using the atmospheric scattering model inversion. The specific steps are as follows:

[0127] S6.1. Using the atmospheric scattering model formula, the fog-free image is inverted. The formula is as follows:

[0128] I(x)=J(x)*t(x)+A[1-t(x)]

[0129] Among them, I(x) represents the foggy image, J(x) refers to the fog-free image, t(x) refers to the transmittance after weighted fusion, and A refers to the atmospheric light value, that is, the global atmospheric light intensity; the inversion formula is:

[0130]

[0131] The atmospheric light value A has been obtained in step S3.2, and the transmittance t(x) after weighted fusion is obtained in step S4, so the fog-free image can be obtained;

[0132] S6.2. Display the fog-free video stream on the system interface.

[0133] like Figure 2 As shown, this embodiment also provides a monitoring image real-time defogging system based on an airport pavement FOD monitoring system, which is used to implement the steps of the above-mentioned monitoring image real-time defogging method based on an airport pavement FOD monitoring system, including:

[0134] Video stream reading module, used to read the airport runway monitoring video stream from the airport pavement FOD monitoring system;

[0135] A dark channel map calculation module is used to calculate the dark channel map of the image using different filter windows;

[0136] A rough transmittance calculation module is used to calculate the rough transmittance obtained by different filter windows;

[0137] Transmittance weighted fusion module, used for weighted fusion of different rough transmittances;

[0138] A transmittance refinement module, used to refine the transmittance after weighted fusion using guided filtering;

[0139] The model inversion module is used to obtain the dehazed image through the inversion of the atmospheric scattering model.

[0140] like Figure 3 As shown, this embodiment also provides a monitoring image real-time defogging operation platform device, which includes a processor, a memory, and a computer program stored in the memory and running on the processor.

[0141] The processor includes one or more processing cores, and the processor is connected to the memory through a bus. The memory is used to store program instructions. When the processor executes the program instructions in the memory, the steps of the above-mentioned real-time defogging system and method for monitoring images based on the airport pavement FOD monitoring system are implemented.

[0142] Optionally, the memory can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk or optical disk.

[0143] In addition, the present invention also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it implements the steps of the above-mentioned real-time defogging system and method for monitoring images based on the airport pavement FOD monitoring system.

[0144] Optionally, the present invention also provides a computer program product comprising instructions, which, when executed on a computer, enables the computer to execute the steps of the above-mentioned aspects of the real-time defogging system and method for monitoring images based on an airport pavement FOD monitoring system.

[0145] A person of ordinary skill in the art can understand that the process of implementing all or part of the steps of the above-mentioned embodiments can be completed by hardware, or can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium, and the above-mentioned storage medium can be a read-only memory, a disk or an optical disk, etc.

[0146] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. The above embodiments and descriptions are only preferred examples of the present invention and are not intended to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A real-time defogging method for monitoring images based on an airport pavement FOD monitoring system, characterized in that: The steps include: S1, read the airport runway monitoring video stream; S2, using different filter windows to calculate the dark channel map of the image; S3, calculating the rough transmittance obtained by different filter windows; S4, performing weighted fusion on the different rough transmittances obtained; S5, using guided filtering to refine the transmittance after weighted fusion; S6. The dehazed image is obtained by inverting the atmospheric scattering model.

2. The real-time defogging method for monitoring images based on the airport pavement FOD monitoring system according to claim 1 is characterized by: In step S1, the specific steps of reading the airport runway monitoring video stream are as follows: S1.

1. Establish connection: obtain the storage location or network address of the monitoring video stream in the airport pavement FOD monitoring system; use the corresponding interface or library to establish a connection with the video stream data source; S1.

2. Initialize the video stream reader: create a video stream reader object according to the selected video processing library or framework; set the relevant parameters of the video stream; S1.3, start reading video frames: enter the loop and read the video stream frame by frame; Use the video stream reader method to obtain the image data of the current frame; perform preliminary processing on the read image data; S1.4, frame data processing: perform necessary preprocessing on each frame of image; the frame data can be stored in a buffer or data structure for subsequent analysis and processing; S1.5, Synchronization and exception handling: handle the synchronization of video streams, ensure the accuracy of frame order and timestamps; monitor possible anomalies during the reading process and take appropriate error handling measures; S1.6, continue reading until the end: continue to loop and read the video frames until the end mark of the video stream is reached or the condition for stopping reading is met; S1.

7. Resource release: After completing the reading of the video stream, release the resources related to the video stream in time to avoid resource leakage.

3. The real-time defogging method for monitoring images based on the airport pavement FOD monitoring system according to claim 1 is characterized by: In step S2, the specific steps of using different filter windows to calculate the dark channel map of the image are as follows: S2.1, read the video stream, and process the images in the video stream frame by frame; S2.2, traverse the single frame image and convert the image into grayscale, and the grayscale value of each pixel is equal to the minimum value of the three-channel color values ​​of each pixel R, G, and B; S2.

3. Select filter windows of f1*f1, f2*f2, ..., fn*fn respectively to perform minimum value filtering on the grayscale window obtained in the previous step to obtain the dark channel image of the image corresponding to the filter window.

4. The real-time defogging method for monitoring images based on the airport pavement FOD monitoring system according to claim 3 is characterized by: In step S2.3, the specific steps of performing minimum value filtering on the grayscale window are as follows: S2.3.1, determine the size of the filter window; S2.3.

2. For each pixel in the image, determine the range of its neighborhood with the pixel as the center; S2.3.

3. Find the minimum value among all pixel values ​​in the neighborhood; S2.3.4, assigning the minimum value to the central pixel as the pixel value after filtering; S2.3.

5. Repeat steps S2.3.2 to S2.3.4 to process each pixel in the image until the entire image is filtered.

5. The real-time defogging method for monitoring images based on the airport pavement FOD monitoring system according to claim 4 is characterized by: In step S3, the specific steps of calculating the rough transmittance obtained by different filter windows are as follows: S3.1, traverse the dark channel images of n filter windows to obtain the positions of the pixels with the largest pixel values ​​in the top 5% of the entire image; S3.2, find the pixel points at the corresponding positions in the first step on the foggy image, and calculate the average value of the R, G, and B channel color values ​​of these pixel points as the atmospheric light value A; S3.3, using the following formula to calculate the dark channel image corresponding to the n filter windows obtained in step S2 to obtain a rough transmittance; the calculation formula for the rough transmittance is: in, is the rough transmittance of the dark channel image, w is the transmittance adjustment parameter, A represents the atmospheric light value, c refers to different color channels, I c (y) represents the color value of color channel c of the foggy image at position y, A c represents the color value of atmospheric light in color channel c, and Ω(x) represents the neighborhood centered on pixel x.

6. The real-time defogging method for monitoring images based on the airport pavement FOD monitoring system according to claim 5 is characterized by: In step S4, the specific steps of weighted fusion of the obtained different rough transmittances are as follows: The rough transmittance obtained in step S3 is weighted fused using the following formula: t(x)=d1t f1 (x)+d2t f2 (x)+...+d n t fn (x) Among them, t(x) is the transmittance after weighted fusion, t f1 (x) represents the transmittance obtained using the f1*f1 window, t f2 (x) represents the transmittance obtained using the f2*f2 window, t fn (x) represents the transmittance obtained using the fn*fn window; d1, d2...d n t f1 (x), t f2 (x)...t fn (x) corresponding weight, and d1+d2+...+d n =1.

7. The real-time defogging method for monitoring images based on the airport pavement FOD monitoring system according to claim 6 is characterized by: In step S5, the specific steps of using guided filtering to refine the transmittance after weighted fusion are as follows: S5.

1. The transmittance map is used as the input image, and the grayscale image of the foggy image is used as the guide image of the guided filtering. The refined transmittance map is obtained by guided filtering, where the formula is as follows: Where O represents the output image, G represents the guide image, i represents the index of the pixel, and w k represents the local window centered at pixel k, a k and b k Represents the coefficient of the linear function when the local window is located at k; Among them, a k and b k The solution is as follows: in, Represents images G and I in local window w k The covariance within Represents the image G in the local window w k The variance within Represents image I in local window w k The mean value in Represents the mean of image G in the local window.

8. The real-time defogging method for monitoring images based on the airport pavement FOD monitoring system according to claim 7 is characterized by: In the step S5.1, the transmittance map as the input image comes from the step S4, and the final transmittance map is obtained by weighted fusion of different rough transmittances processed in the step S3; the grayscale map of the foggy image as the guide image comes from the step S2.2, in which the single frame image is traversed and grayscaled, and the grayscale map of the foggy image is selected.

9. The real-time defogging method for monitoring images based on the airport pavement FOD monitoring system according to claim 8 is characterized by: In step S6, the specific steps of obtaining the defogging image by inverting the atmospheric scattering model are as follows: S6.

1. Using the atmospheric scattering model formula, the fog-free image is inverted. The formula is as follows: I(x)=J(x)*t(x)+A[1-t(x)] Among them, I(x) represents the foggy image, J(x) refers to the fog-free image, t(x) refers to the transmittance after weighted fusion, and A refers to the atmospheric light value, that is, the global atmospheric light intensity; the inversion formula is: The atmospheric light value A has been obtained in step S3.2, and the transmittance t(x) after weighted fusion is obtained in step S4, so the fog-free image can be obtained; S6.

2. Display the fog-free video stream on the system interface.

10. A monitoring image real-time defogging system based on an airport pavement FOD monitoring system, used to implement the steps of the monitoring image real-time defogging method based on an airport pavement FOD monitoring system according to any one of claims 1 to 9, characterized in that: include: Video stream reading module, used to read the airport runway monitoring video stream from the airport pavement FOD monitoring system; A dark channel map calculation module is used to calculate the dark channel map of the image using different filter windows; A rough transmittance calculation module is used to calculate the rough transmittance obtained by different filter windows; Transmittance weighted fusion module, used for weighted fusion of different rough transmittances; A transmittance refinement module, used to refine the transmittance after weighted fusion using guided filtering; The model inversion module is used to obtain the dehazed image through the inversion of the atmospheric scattering model.