An assembly workshop behavior monitoring system

Abstract

The present invention belongs to the field of workshop monitoring, and discloses an assembly workshop behavior monitoring system, including a shooting device, an analysis device and a prompting device; the shooting device is used to shoot the assembly workshop to obtain a monitoring image; the analysis device is used to cut the monitoring image to obtain the worker area in the monitoring image, and is used to identify the worker area to detect whether the worker is wearing safety protection equipment correctly; the prompting device is used to issue a prompt to the safety officer of the assembly workshop when it is detected that there are workers who are not wearing safety protection equipment correctly. The present invention can more comprehensively and more timely detect whether the workers in the assembly workshop are wearing safety protection equipment correctly, thereby greatly reducing the probability of the occurrence of the event that the worker who is not wearing safety protection equipment correctly can be found in time.

Description

Technical Field

[0001] The present invention relates to the field of workshop monitoring, and in particular to an assembly workshop behavior monitoring system. Background Art

[0002] An assembly workshop is a workplace used to produce and assemble various products. It is the place where one of the processes in the overall process of industrial production is realized. The assembly workshop includes a series of production equipment and workbenches, where workers assemble parts into complete products according to a certain process flow.

[0003] During the assembly process, workshop workers are usually required to wear safety protection equipment such as hard hats and goggles. However, some workers, due to laziness or carelessness, always fail to wear safety protection equipment correctly, which obviously makes workers vulnerable to injury, and the company also needs to compensate the injured workers. Most assembly workshops will set up the position of safety officer to urge workers to wear protective equipment correctly. However, safety officers cannot always observe all workers in real time, so there is still a high probability that workers who do not wear safety protection equipment correctly will not be discovered in time. Summary of the invention

[0004] The purpose of the present invention is to disclose an assembly workshop behavior monitoring system to solve the technical problems raised in the background technology.

[0005] In order to achieve the above object, the present invention adopts the following technical solution:

[0006] The present invention provides an assembly workshop behavior monitoring system, including a shooting device, an analyzing device and a prompting device;

[0007] The photographing device is used to photograph the assembly workshop to obtain monitoring images;

[0008] The analysis device is used to cut the monitoring image, obtain the worker area in the monitoring image, and to identify the worker area to detect whether the worker is wearing the safety protection equipment correctly;

[0009] The reminder device is used to remind the safety officer in the assembly workshop when it detects the presence of workers who are not wearing safety protection equipment correctly.

[0010] Preferably, the photographing device includes a photographing planning device and a photographing device;

[0011] The camera is installed on the wall of the assembly workshop, and the camera lens is tilted downward to obtain monitoring images;

[0012] There are multiple shooting devices, and each of the multiple shooting devices shoots different areas of the assembly workshop;

[0013] The shooting planning device is used to determine the shooting cycle of each shooting device respectively;

[0014] The photographing device is used to photograph the assembly workshop based on a photographing cycle to obtain monitoring images.

[0015] Preferably, the prompting device includes a positioning module and a prompting module;

[0016] The positioning module is used to obtain the installation position of the shooting device corresponding to the monitoring image when the analysis device detects that the worker does not wear the safety protection equipment correctly based on the monitoring image;

[0017] The prompt module is used to display the installation location of the shooting device corresponding to the monitoring image.

[0018] Preferably, the analysis device comprises a cutting module and a recognition module;

[0019] The cutting module is used to cut the monitoring image to obtain the worker area in the monitoring image, including:

[0020] Perform differential processing on the monitoring image and the background image to obtain a differential area;

[0021] Acquire the worker area in the monitoring image based on the differential area;

[0022] The identification module is used to identify the worker area and detect whether the workers are wearing safety protection equipment correctly, including:

[0023] The worker area is input into the trained detection model for detection to check whether the workers are wearing safety protection equipment correctly.

[0024] Preferably, the surveillance image and the background image have the same shooting angle, the same light intensity and the same focal length as the shooting environment;

[0025] The background image is an image obtained by the camera photographing the assembly when there are no workers in the assembly workshop.

[0026] Preferably, performing differential processing on the monitoring image and the background image to obtain a differential area includes:

[0027] Grayscale the monitoring image and background image respectively to obtain images MP and BP;

[0028] Calculate the grayscale change value of each pixel in MP respectively;

[0029] Get all connected domains consisting of pixels in MP whose grayscale change value is greater than the set grayscale value, and take each connected domain as a differential region.

[0030] Preferably, the grayscale change value is calculated as:

[0031] dif s = | gray s,MP -gray s,BP |

[0032] dif s Indicates the grayscale change value of pixel s in MP, gray s,MP Represents the gray value of pixel s in MP; gray s,BP Indicates the grayscale value of the pixel in BP that has the same coordinates as MP.

[0033] Preferably, acquiring the worker area in the monitoring image based on the differential area includes:

[0034] Performing repair processing on each differential region respectively to obtain a repaired differential region;

[0035] Calculate the aspect ratio of each repaired difference region respectively;

[0036] The repaired difference regions whose aspect ratio is smaller than the set aspect ratio threshold are deleted, and the remaining repaired difference regions are used as worker regions.

[0037] Preferably, the process of repairing the differential area is as follows:

[0038] Establish a rectangular coordinate system in MP;

[0039] For the difference area q, the minimum X-axis coordinate cx, the maximum X-axis coordinate mx, the minimum Y-axis coordinate cy and the maximum Y-axis coordinate my of the pixel points in the difference area q are obtained;

[0040] Calculate the restoration coefficients of the pixels in region Q. The X-axis coordinates x and Y-axis coordinates y of the pixels in region Q satisfy the following conditions:

[0041] cx≤x≤mx and cy≤y≤my;

[0042] The differential region is repaired based on the repair coefficient to obtain a repaired differential region.

[0043] Preferably, the calculation formula of the repair coefficient is:

[0044]

[0045] repr q represents the repair coefficient of the differential area q, gray urepresents the gray value of pixel u in region Q, NQ represents the total number of pixels in region Q, aveQ represents the average gray value of pixels in region Q, MT represents the number of differential regions in set Uq, and Uq represents the distance from differential region q less than MF represents the number of differential regions in Uq whose average grayscale value between q and q is less than the set average grayscale value threshold, and λ represents the calculation weight.

[0046] Beneficial effects:

[0047] The present invention obtains monitoring images of different areas in the assembly workshop by installing a camera in the assembly workshop, and also uses an analysis device to automatically identify the monitoring images, so that when an event that a worker does not wear safety protection equipment correctly is detected, the safety officer can be notified in a timely manner. Compared with the method of simply relying on the safety officer to observe, the present invention can more comprehensively and timely detect whether the workers in the assembly workshop are wearing safety protection equipment correctly, thereby greatly reducing the probability of timely discovering workers who do not wear safety protection equipment correctly. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0049] Figure 1 A schematic diagram of an assembly workshop behavior monitoring system of the present invention.

[0050] Figure 2 The figure is a schematic diagram of a process of acquiring a worker region in a monitoring image based on a differential region according to the present invention. DETAILED DESCRIPTION

[0051] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. The components of the embodiments of the present invention generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed invention, but merely represents the selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making creative work belong to the scope of protection of the present invention.

[0052] like Figure 1 In one embodiment shown, the present invention provides an assembly workshop behavior monitoring system, including a photographing device, an analyzing device and a prompting device;

[0053] The photographing device is used to photograph the assembly workshop to obtain monitoring images;

[0054] The analysis device is used to cut the monitoring image, obtain the worker area in the monitoring image, and to identify the worker area to detect whether the worker is wearing the safety protection equipment correctly;

[0055] The reminder device is used to remind the safety officer in the assembly workshop when it detects the presence of workers who are not wearing safety protection equipment correctly.

[0056] The above scheme installs a camera in the assembly workshop to obtain monitoring images of different areas in the assembly workshop, and also uses an analysis device to automatically identify the monitoring images, so that when an event that a worker does not wear safety protection equipment correctly is detected, the safety officer can be notified in a timely manner. Compared with the method of relying solely on observation by the safety officer, the present invention can more comprehensively and timely detect whether the workers in the assembly workshop are wearing safety protection equipment correctly, thereby greatly reducing the probability of timely discovering workers who do not wear safety protection equipment correctly.

[0057] Preferably, the photographing device includes a photographing planning device and a photographing device;

[0058] The camera is installed on the wall of the assembly workshop, and the camera lens is tilted downward to obtain monitoring images;

[0059] There are multiple shooting devices, and each of the multiple shooting devices shoots different areas of the assembly workshop;

[0060] The shooting planning device is used to determine the shooting cycle of each shooting device respectively;

[0061] The photographing device is used to photograph the assembly workshop based on a photographing cycle to obtain monitoring images.

[0062] Specifically, the tilted downward shooting may be shooting at a tilt of 45° downward.

[0063] Preferably, determining the shooting period of each shooting device separately includes:

[0064] The shooting cycle of each shooting device is determined in the following manner:

[0065] Determine whether the current time is during working hours or non-working hours;

[0066] If it is a non-working time period, the shooting cycle is set to T1, and if it is a working time period, the shooting cycle is set to T2, and T2 is smaller than T1.

[0067] By setting different shooting cycles for different time periods, it is possible to avoid shooting too many monitoring images during non-working time periods and extend the service life of the shooting equipment. Since there are no workers during non-working time periods, a longer shooting cycle can be set.

[0068] During non-working hours, the camera can be used as a monitoring device to shoot the assembly workshop.

[0069] Preferably, T2 is 30 seconds.

[0070] Furthermore, T1 may be 90 seconds.

[0071] Preferably, photographing the assembly workshop based on a photographing cycle to obtain monitoring images includes:

[0072] The assembly workshop is photographed once every photographing cycle to obtain monitoring images, and the obtained monitoring images are transmitted to the analysis device.

[0073] Preferably, the prompting device includes a positioning module and a prompting module;

[0074] The positioning module is used to obtain the installation position of the shooting device corresponding to the monitoring image when the analysis device detects that the worker does not wear the safety protection equipment correctly based on the monitoring image;

[0075] The prompt module is used to display the installation location of the shooting device corresponding to the monitoring image.

[0076] Specifically, the installation location of each shooting device is pre-stored in the positioning module. The name of the shooting device can be obtained from the Exif information of the monitoring image. Different names can be set for shooting devices in different installation locations, so that the corresponding shooting device can be determined based on the Exif information of the monitoring image, and then the installation location of the shooting device can be determined.

[0077] After the prompt module displays the installation location, the safety officer can go to that location, observe the workers around that location, and remind the workers who do not wear safety protection equipment correctly.

[0078] Preferably, the analysis device comprises a cutting module and a recognition module;

[0079] The cutting module is used to cut the monitoring image to obtain the worker area in the monitoring image, including:

[0080] Perform differential processing on the monitoring image and the background image to obtain a differential area;

[0081] Acquire the worker area in the monitoring image based on the differential area;

[0082] The identification module is used to identify the worker area and detect whether the workers are wearing safety protection equipment correctly, including:

[0083] The worker area is input into the trained detection model for detection to check whether the workers are wearing safety protection equipment correctly.

[0084] Preferably, the detection model can be a CNN model.

[0085] The training process of the detection model is as follows:

[0086] Data collection: Collect images of different workers wearing safety protective equipment correctly and not wearing safety protective equipment correctly in different scenarios.

[0087] Data annotation: Annotate the collected images to indicate whether the workers in the images are wearing safety protection equipment correctly.

[0088] Model training: Use the labeled images to train the CNN model so that the trained CNN model can be used to determine whether the workers in the image have correctly worn safety protection equipment.

[0089] Specifically, if the worker does not wear all types of safety protection equipment, it means that the worker is not wearing the safety protection equipment correctly.

[0090] Preferably, the surveillance image and the background image have the same shooting angle, the same light intensity and the same focal length as the shooting environment;

[0091] The background image is an image obtained by the camera photographing the assembly when there are no workers in the assembly workshop.

[0092] Since there are no workers in the background image, the area where the workers exist can be quickly determined by image difference, thereby improving the speed of obtaining the detection result of the present invention.

[0093] Shooting angle refers to the pitch angle of the lens.

[0094] Preferably, performing differential processing on the monitoring image and the background image to obtain a differential area includes:

[0095] Grayscale the monitoring image and background image respectively to obtain images MP and BP;

[0096] Calculate the grayscale change value of each pixel in MP respectively;

[0097] Get all connected domains consisting of pixels in MP whose grayscale change value is greater than the set grayscale value, and take each connected domain as a differential region.

[0098] Since the shooting angle, the illumination intensity of the shooting environment and the focal length are the same, when the grayscale value of the pixels in a certain area changes greatly compared with the pixels in the background image, it can be considered that there are workers in the area.

[0099] Specifically, the grayscale value is set to 10.

[0100] Preferably, the grayscale change value is calculated as:

[0101] dif s = | gray s,MP -gray s,BP |

[0102] dif s Indicates the grayscale change value of pixel s in MP, gray s,MP Represents the gray value of pixel s in MP; gray s,BP Indicates the grayscale value of the pixel in BP that has the same coordinates as MP.

[0103] Specifically, during calculation, BM and MP are placed in the same coordinate system.

[0104] Preferably, if Figure 2 As shown, the worker area in the monitoring image is obtained based on the differential area, including:

[0105] Performing repair processing on each differential region respectively to obtain a repaired differential region;

[0106] Calculate the aspect ratio of each repaired difference region respectively;

[0107] The repaired difference regions whose aspect ratio is smaller than the set aspect ratio threshold are deleted, and the remaining repaired difference regions are used as worker regions.

[0108] Specifically, the aspect ratio threshold is set to 6.

[0109] Specifically, the aspect ratio of the repaired differential region is calculated as follows:

[0110] Get the difference Z1 between the maximum and minimum values ​​of the X-axis coordinates of the pixel points in the repaired differential area, and get the difference Z2 between the maximum and minimum values ​​of the Y-axis coordinates of the pixel points in the repaired differential area, then the aspect ratio is

[0111] Specifically, the X-axis coordinate is parallel to the direction of the rows of the pixels of the MP, and the Y-axis coordinate is parallel to the direction of the columns of the pixels of the MP.

[0112] Specifically, each worker area corresponds to one worker.

[0113] Preferably, the process of repairing the differential area is as follows:

[0114] Establish a rectangular coordinate system in MP;

[0115] For the difference area q, the minimum X-axis coordinate cx, the maximum X-axis coordinate mx, the minimum Y-axis coordinate cy and the maximum Y-axis coordinate my of the pixel points in the difference area q are obtained;

[0116] Calculate the restoration coefficients of the pixels in region Q. The X-axis coordinates x and Y-axis coordinates y of the pixels in region Q satisfy the following conditions:

[0117] cx≤x≤mx and cy≤y≤my;

[0118] The differential region is repaired based on the repair coefficient to obtain a repaired differential region.

[0119] The reason for repairing is that if the grayscale value difference between some pixels in the worker area and the pixels in the background area of ​​the monitoring image is too small, the worker area may be incomplete, thus affecting the accuracy of the final detection result. Repairing is conducive to obtaining a more complete worker area.

[0120] Preferably, the calculation formula of the repair coefficient is:

[0121]

[0122] repr q represents the repair coefficient of the differential area q, gray u represents the gray value of pixel u in region Q, NQ represents the total number of pixels in region Q, aveQ represents the average gray value of pixels in region Q, MT represents the number of differential regions in set Uq, and Uq represents the distance from differential region q less than MF represents the number of differential regions in Uq whose average grayscale value between q and q is less than the set average grayscale value threshold, and λ represents the calculation weight.

[0123] The repair coefficient of the present invention is calculated based on the fluctuation degree of the grayscale value of the pixel points in Q and the number of differential areas around q that are close to the average grayscale value of q. Therefore, when the fluctuation degree is smaller and the number of differential areas around q that are close to the average grayscale value of q is smaller, the repair coefficient is smaller, and at this time, the probability that q and the surrounding differential areas belong to the same complete area is lower; when the fluctuation degree is larger and the number of differential areas around q that are close to the average grayscale value of q is larger, the repair coefficient is larger, and at this time, the probability that q and the surrounding differential areas belong to the same complete area is higher. In this way, the present invention can subsequently use different repair methods to repair q based on the repair coefficient, so as to achieve a better balance between ensuring the accuracy of the repair and the efficiency of the repair.

[0124] Specifically, the average grayscale value is the average grayscale value of the pixels in the differential area.

[0125] Specifically, the calculation weight is 0.5.

[0126] Preferably, the distance between two differential regions is the distance between the centers of the two differential regions.

[0127] Preferably, performing repair processing on the differential region based on the repair coefficient to obtain a repaired differential region includes:

[0128] For the difference region q, if repr q If it is greater than the set repair coefficient threshold, the first repair method is used to repair the differential area q to obtain a repaired differential area; otherwise, the second repair method is used to repair the differential area to obtain a repaired differential area.

[0129] Specifically, the repair coefficient threshold is 0.3.

[0130] Preferably, the first repairing method is used to repair the differential region q to obtain a repaired differential region, including:

[0131] The first step is to store the pixels of the differential area q into the set PU1;

[0132] The second step is to obtain the set PU2 of edge pixels in PU1;

[0133] The third step is to perform region growing processing with each pixel in PU2 as a seed pixel within the range of region Q, and obtain the connected domain generated by each pixel in PU2 after the region growing is completed;

[0134] Step 4: Obtain a set PU3 of pixels in all connected domains obtained in step 3;

[0135] In the fifth step, the connected domain formed by the union of PU3 and PU1 is used as the repaired differential region.

[0136] Specifically, when the repair coefficient is relatively large, the present invention performs repair processing based on the pixel points in PU2 respectively, and performs repair at the pixel level, so the repair result is more accurate.

[0137] Specifically, in the region growing process of the third step, the growing threshold is 5, that is, if the absolute value of the difference in grayscale value between the pixel point in the domain and the seed pixel point is less than or equal to 5, the pixel point is included in the connected domain where the seed pixel point is located.

[0138] Preferably, the second repairing method is used to repair the differential region q to obtain a repaired differential region, including:

[0139] For region Q, a 2×2 window is used to slide non-overlappingly in region Q to obtain multiple local regions;

[0140] The following calculations are performed for each local area:

[0141] Step 1: for a local area s, determine whether all pixels in s belong to set PU1. If so, store the pixels in s into set PU3; if not, proceed to step 2.

[0142] Step 2, taking the local area s as the target area;

[0143] Step 3, obtain a set su of local areas adjacent to the target area and including pixels that do not belong to the set PU1 or PU3, and determine whether su is an empty set. If so, end the calculation, otherwise, proceed to step 4;

[0144] Step 4, respectively calculate the difference between the average gray value of each local area in su and s;

[0145] Step 5, determining whether the minimum value of the difference obtained in step 5 is less than the set difference threshold, if so, storing the pixel points in the local area corresponding to the minimum difference obtained in step 4 into set PU3;

[0146] Step 6: Take the local area corresponding to the minimum difference obtained in step 4 as the next target area and go to step 3.

[0147] When the repair coefficient is small, the present invention performs repair processing at the regional level, which can greatly improve the efficiency of the repair process. Each calculation is performed on multiple pixels, so the efficiency of the repair process can be greatly improved. Moreover, since the repair coefficient is small, the probability that the gray value of the pixels in the local area fluctuates less is greater. Therefore, even if the repair processing is performed at the regional level, the difference between the repair result and the repair processing at the pixel level is small, so that the repair efficiency can be improved while ensuring the repair effect.

[0148] Specifically, the difference between the average grayscale values ​​of each local area in su and s is |B1-B2|, where B1 is the average grayscale value of the pixels in the local area, and B2 is the average grayscale value of the pixels in s.

[0149] Specifically, the difference threshold is set to 5.

[0150] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. An assembly workshop behavior monitoring system, characterized in that: It includes a shooting device, an analyzing device and a prompting device; The photographing device is used to photograph the assembly workshop to obtain monitoring images; The analysis device is used to cut the monitoring image, obtain the worker area in the monitoring image, and to identify the worker area to detect whether the worker is wearing the safety protection equipment correctly; The reminder device is used to remind the safety officer in the assembly workshop when it detects the presence of workers who are not wearing safety protection equipment correctly.

2. The assembly workshop behavior monitoring system according to claim 1, characterized in that: The shooting device includes shooting planning equipment and shooting equipment; The camera is installed on the wall of the assembly workshop, and the camera lens is tilted downward to obtain monitoring images; There are multiple shooting devices, and each of the multiple shooting devices shoots different areas of the assembly workshop; The shooting planning device is used to determine the shooting cycle of each shooting device respectively; The photographing device is used to photograph the assembly workshop based on a photographing cycle to obtain monitoring images.

3. The assembly workshop behavior monitoring system according to claim 2, characterized in that: The prompting device includes a positioning module and a prompting module; The positioning module is used to obtain the installation position of the shooting device corresponding to the monitoring image when the analysis device detects that the worker does not wear the safety protection equipment correctly based on the monitoring image; The prompt module is used to display the installation location of the shooting device corresponding to the monitoring image.

4. The assembly workshop behavior monitoring system according to claim 1, characterized in that: The analysis device includes a cutting module and a recognition module; The cutting module is used to cut the monitoring image to obtain the worker area in the monitoring image, including: Perform differential processing on the monitoring image and the background image to obtain a differential area; Acquire the worker area in the monitoring image based on the differential area; The identification module is used to identify the worker area and detect whether the workers are wearing safety protection equipment correctly, including: The worker area is input into the trained detection model for detection to check whether the workers are wearing safety protection equipment correctly.

5. The assembly workshop behavior monitoring system according to claim 4, characterized in that: The surveillance image and the background image have the same shooting angle, lighting intensity and focal length; The background image is an image obtained by the camera photographing the assembly when there are no workers in the assembly workshop.

6. The assembly workshop behavior monitoring system according to claim 5, characterized in that: Perform differential processing on the monitoring image and the background image to obtain the differential area, including: Grayscale the monitoring image and background image respectively to obtain images MP and BP; Calculate the grayscale change value of each pixel in MP respectively; Get all connected domains consisting of pixels in MP whose grayscale change value is greater than the set grayscale value, and take each connected domain as a differential region.

7. The assembly workshop behavior monitoring system according to claim 6, characterized in that: The calculation formula of grayscale change value is: dif s =|gray s,MP -gray s,BP | dif s Indicates the grayscale change value of pixel s in MP, gray s,MP Represents the gray value of pixel s in MP; gray s , BP Indicates the grayscale value of the pixel in BP that has the same coordinates as MP.

8. The assembly workshop behavior monitoring system according to claim 6, characterized in that: The worker area in the monitoring image is obtained based on the difference area, including: Performing repair processing on each differential region respectively to obtain a repaired differential region; Calculate the aspect ratio of each repaired difference region respectively; The repaired difference regions whose aspect ratio is smaller than the set aspect ratio threshold are deleted, and the remaining repaired difference regions are used as worker regions.

9. The assembly workshop behavior monitoring system according to claim 6, characterized in that: The process of repairing the differential area is as follows: Establish a rectangular coordinate system in MP; For the difference area q, the minimum X-axis coordinate cx, the maximum X-axis coordinate mx, the minimum Y-axis coordinate cy and the maximum Y-axis coordinate my of the pixel points in the difference area q are obtained; Calculate the restoration coefficients of the pixels in region Q. The X-axis coordinates x and Y-axis coordinates y of the pixels in region Q satisfy the following conditions: cx≤x≤mx and cy≤y≤my; The differential region is repaired based on the repair coefficient to obtain a repaired differential region.

10. The assembly workshop behavior monitoring system according to claim 9, characterized in that: The calculation formula of the repair coefficient is: repr q represents the repair coefficient of the differential area q, gray u represents the gray value of pixel u in region Q, NQ represents the total number of pixels in region Q, aveQ represents the average gray value of pixels in region Q, MT represents the number of differential regions in set Uq, and Uq represents the distance between differential region q and the pixel u. MF represents the number of differential regions in Uq whose average grayscale value between q and q is less than the set average grayscale value threshold, and λ represents the calculation weight.

Images