Weak and small target detection method and system based on polarization consistency local contrast

Abstract

The invention discloses a weak and small target detection method and system based on polarization consistency local contrast, and the method comprises the steps: obtaining a gray image of a weak and small target and a plurality of corresponding polarization images; realizing fusion noise reduction of the grayscale image and the polarization image based on a guided filtering algorithm to obtain an output image; traversing the whole output image by utilizing a mechanism of local contrast and combining the combined polarization enhancement coefficient to obtain a polarization degree saliency map of the whole image; in the polarization degree saliency map, a self-adaptive threshold value is used for segmenting a dark and weak target and a background, and weak and small target detection is completed. By utilizing the method, the difficulty in detecting the infrared weak and small target under the complex background can be solved, and the accurate detection of the infrared weak and small target is realized.

Description

Technical Field

[0001] The present invention belongs to the field of photoelectric detection and imaging technology, and in particular relates to a method and system for detecting small targets based on polarization consistency local contrast. Background Art

[0002] Enhanced detection of small targets is one of the key technologies in application fields such as guidance, early warning, airborne or satellite-borne monitoring, and anti-UAV, and is also an important research direction of traditional small target detection.

[0003] However, traditional weak target detection can only obtain targets in a specific background, that is, by extracting the difference information between each pixel in the image and its neighboring pixels through the obvious grayscale difference between the background grayscale and the target grayscale, the target can be successfully screened out.

[0004] For example, a Chinese patent document with publication number CN112396061A discloses an Otsu target detection method based on target grayscale tendency weighting, which includes the following steps: Step S1: grayscale the image to be tested to obtain the grayscale histogram of the image to be tested; Step S2: determine the expected threshold value according to the obtained grayscale histogram of the image to be tested; Step S3: calculate the occurrence probability of foreground and background pixels and their grayscale mean according to the expected threshold value; Step S4: select the grayscale tendency coefficient to be changed according to the expected threshold value, if the expected threshold value is closer to the high grayscale area, select the high tendency coefficient α, otherwise if it is closer to the low grayscale area, select the low tendency coefficient β; Step S5: use the improved Otsu inter-class variance formula to weight the gradient valley value, obtain the threshold value that maximizes the variance, and realize the separation of the target and the background.

[0005] A Chinese patent document with publication number CN117422740A discloses an infrared-based moving target detection and processing method, comprising: taking an infrared moving target image, first filtering the image; then segmenting the target area and removing the shadow using a method based on grayscale value changes; binarizing and dilating the segmented target area; then searching for contours; and after the target contour screening operation, identifying and locking the position of the infrared target in the image.

[0006] However, when faced with various complex scenes such as complex weather such as rain, snow, haze, and blowing sand, scattering media such as smoke and underwater, or non-cooperative targets such as camouflaged targets and transparent objects, these traditional weak target detection methods will lose their detection capabilities.

[0007] To solve the above problems, the local contrast method based on the characteristics of the human visual system is currently proposed. That is, according to the characteristics of human vision, the salient area of ​​the target is quickly obtained from the complex background and the invalid background information is eliminated.

[0008] However, the local contrast method faces the following two difficulties: First, due to the limitations of the definition of the local contrast principle, only the difference information between the target and the neighborhood background is extracted during the calculation process, and no enhancement operation is performed on the real target. When the target is very weak, it is easy to cause missed detection; secondly, the background is not subtracted during the calculation. When there is a large area of ​​high-brightness background area in the image, it is easy to cause false detection when the image is enhanced. Summary of the invention

[0009] The present invention provides a small target detection method and system based on polarization consistency local contrast, which can solve the problem of detecting small infrared targets under complex backgrounds and realize accurate detection of small infrared targets.

[0010] The small target detection method based on polarization consistency local contrast includes the following steps: (1) Obtain the grayscale image of the dim target and the corresponding multiple polarization images; (2) Based on the guided filtering algorithm, the grayscale image and polarization image are fused and denoised to obtain the output image; (3) Using the mechanism of local contrast and the joint polarization enhancement coefficient, the entire output image is traversed to obtain the polarization saliency map of the entire image; (4) In the polarization saliency map, an adaptive threshold is used to segment the dim target from the background to complete the dim target detection.

[0011] In step (1), the polarizer is rotated to change the angle of the analyzer. The rotation angles are 0°, 45°, 90°, and 135°, respectively. The corresponding four polarization images are: , , , .

[0012] The specific process of step (2) is as follows: The grayscale image is used as the guide image and the polarization image is used as the image to be filtered for guided filtering; for the non-target area, the gradient of the grayscale image is small, the output polarization image is the result of smoothing, and the noise is suppressed; for the target area, the gradient of the grayscale image is large, and the target is retained in the output polarization image.

[0013] The specific process of step (3) is as follows: Slide the sliding window from left to right and from top to bottom to traverse the entire image area and divide it into several sliding window areas; Each sliding window area is regarded as a sub-area, and each sub-area is divided into a central pixel area. and surrounding pixel areas; get the central pixel area Medium polarization maximum , mean polarization degree And the average polarization value of the surrounding pixel area , Indicates the number of the surrounding pixel area; Constructing joint polarization enhancement coefficient based on degree of polarization and polarization consistency ; Introducing the joint polarization enhancement factor The polarization significance map is: ; In the formula, Indicates the sequence number of the sliding window area, Indicates The polarization saliency map of the joint polarization enhancement coefficient is introduced in each sliding window region. Indicates the center pixel area and the The grayscale difference ratio of the surrounding pixel areas.

[0014] The entire image is traversed to obtain the polarization saliency map of the entire image.

[0015] Constructing joint polarization enhancement coefficient based on degree of polarization and polarization consistency , the formula is as follows: ; In the formula, is the polarization degree in the entire sliding window area; is the polarization standard deviation in the entire sliding window area, which represents the polarization consistency.

[0016] and The calculation formula is as follows: ; ; In the formula, Indicates light intensity information, and Represents linear polarization information, , and Obtained by the following formula: ; In the formula, Represents the value obtained by multiple polarization images Vector, Indicates light intensity information, and Represents linear polarization information, Represents circular polarization information. Since the ground-based polarization detection system does not collect circular polarization information, The weight is 0; , , , These are the four polarization images corresponding to the analysis angles of the rotating polarizer being 0°, 45°, 90°, and 135°.

[0017] In step (5), the adaptive threshold Defined as: ; in, and are the mean and variance of the polarization saliency map, is an adjustable constant parameter. Value greater than threshold When , the pixel is the target, otherwise it is the background.

[0018] The present invention also provides a small target detection system based on polarization consistency local contrast, which is characterized in that it includes a memory and one or more processors, wherein the memory stores executable code, and when the one or more processors execute the executable code, they are used to implement the above-mentioned small target detection method.

[0019] Compared with the prior art, the present invention has the following beneficial effects: The present invention obtains polarization images and grayscale images of small targets through a polarization detection system, fuses polarization images and grayscale images based on a guided filtering algorithm to achieve polarization image noise reduction, obtains a joint polarization coefficient through the characteristics of polarization degree and polarization direction consistency; enhances the polarization image target through the target contrast of the joint polarization coefficient; and realizes the detection of small targets through an adaptive threshold, thereby finally achieving enhanced detection of infrared small targets. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 This is a flow chart of the dim target detection method based on polarization consistency local contrast of the present invention.

[0021] Figure 2 Schematic diagram of calculating a polarization saliency map in an embodiment of the present invention. DETAILED DESCRIPTION

[0022] The present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be pointed out that the embodiments described below are intended to facilitate the understanding of the present invention and do not have any limiting effect on the present invention.

[0023] The present invention takes into account that polarization information is independent of radiation intensity information, and can highlight high-frequency information such as target edges and details under complex backgrounds. In addition, polarization information is determined by factors such as object material and light irradiation angle. Therefore, it is more suitable for detecting weak targets under complex backgrounds. Based on these characteristics of polarization information, polarization information is applied to the local contrast method, and enhanced detection of weak targets is achieved through the polarization difference between the target and the background.

[0024] like Figure 1 As shown, the small target detection method based on polarization consistency local contrast includes the following steps: Step 1: Obtain a grayscale image and a polarization image of a dim target.

[0025] First, the grayscale image of the weak target is obtained through the polarization detection system, and then the polarizer is rotated to change the angle of the analyzer angle. The rotation angles are 0°, 45°, 90°, and 135° respectively. The corresponding polarization images of the same scene are: , , , .

[0026] Step 2: Implement denoising by fusion of polarization image and grayscale image based on guided filtering algorithm.

[0027] Using a smoothing filter to filter an image can suppress the noise in the image, but this will also reduce the contrast of small targets. Therefore, the ideal processing effect is to smooth the noise in the image without changing the contrast between small targets and the surrounding background.

[0028] Combining the characteristics of polarization image and grayscale image, the guided filtering algorithm is used to fuse the polarization image and grayscale image to achieve polarization image denoising. Guided filtering is an edge-preserving filter that requires the input of a guide image and an image to be filtered. It is then assumed that there is a local linear relationship between the output image and the guide image, that is, the following equation holds: ; in, and Represents the output image and guide image A pixel in , represents the area of ​​the sliding window, Display Window The linear relationship parameters satisfied by the guide image and the output image.

[0029] The guided filtering algorithm defines the input image With the output image The error function is: ; ; in, is the input and output error function, is a pixel in the input image, and the coefficient corresponding to the minimum error function is solved : ; ; in, and Represent the guide images In the window The mean and variance of Representation calculation and The covariance of Indicates that the input image is in the window Finally, the grayscale mean of the window is calculated according to the above formula Any pixel in The output of , and then for all pixels containing The calculation results of the window are averaged to get the output image ; in, , From the derivation, we can know that the guided filter can input different regions of the guided image according to the gradient of the guided image. If the input is a smooth region of the guided image, the calculation is Very small, near , so the output image is the result of smoothing the input image; if the input is a region with a large gradient of the guide image, the guide image has a large variance, and the calculation is Close to 1, is close to 0, so the output image also has a large gradient.

[0030] The grayscale image is used as the guide image and the polarization image is used as the image to be filtered for guided filtering. For non-target areas, the gradient of the grayscale image is small, and the output polarization image is the result of smoothing, and the noise is suppressed; for the target area, the gradient of the grayscale image is large, and the output polarization image target is retained. Therefore, using guided filtering to process polarization images can effectively suppress noise while retaining the polarization image target.

[0031] Step 3: Obtain the joint polarization enhancement coefficient.

[0032] The polarization detection system obtains the grayscale image of the scene by changing the angle of the analyzer. The rotation angles are 0°, 45°, 90°, and 135° respectively. The corresponding collected images are , , , , through these four images, we can calculate Vector: ; In the formula, express Vector, Indicates light intensity information, and Represents linear polarization information, Represents circular polarization information. Since the ground-based polarization detection system does not collect circular polarization information, The weight is 0. Vectors can represent the entire polarization state of light, and often characterize polarization state information through polarization degree and polarization angle. The degree of polarization represents the ratio of the intensity of completely polarized light in a beam to the total light intensity, and the polarization angle represents the direction of polarization of the beam. The calculation method is as follows: ; ; Polarization consistency mainly describes the consistency of polarization direction, which refers to the polarization angle in the detection area. Specifically, the polarization angle in the target area changes less and the polarization direction is more uniform, that is, the target area has higher polarization consistency. The polarization distribution in the background area is more random, resulting in a large change in the polarization angle in the area, that is, the background area has lower polarization consistency. The dual parameters of polarization degree and polarization consistency construct the joint polarization enhancement coefficient. The calculation is as follows: ; Step 4: target contrast enhancement method combining polarization degree and polarization angle.

[0033] According to the analysis of the polarization characteristics of the target and the background, there are obvious differences in the polarization degree and polarization direction consistency between the target and the background. Therefore, it is proposed to use the dual parameters of polarization degree and polarization consistency to detect dim targets. Local contrast is the contrast between the target and the background, which can better highlight the target and suppress the background, making the target more prominent during the detection process. Therefore, using the mechanism of local contrast, a polarization degree saliency map based on the dual-parameter enhancement of polarization degree and polarization direction consistency is proposed.

[0034] The polarization saliency map is obtained by calculating the local contrast, which evaluates the contrast difference between the central pixel and the surrounding pixels through a sliding window. As shown, represents the central pixel area, 1 to 8 represent the surrounding pixel area, and the sliding window is slid from left to right and from top to bottom to traverse the entire image area and divide it into several sliding window areas.

[0035] In each sliding window area, the sub-area is divided into the central pixel area , surrounding pixel area (1~8).

[0036] The maximum polarization value of the central pixel area is: ; in, is the maximum polarization degree in region T, is the polarization degree value of each pixel in area T.

[0037] The mean polarization degree of the sub-region is: ; in, is the mean polarization degree in the sub-region, is the polarization degree value of each pixel in the sub-region. is the mean polarization degree in region T, is the average polarization degree in the surrounding pixel area.

[0038] The polarization degree and polarization direction consistency difference between the central pixel area and the background area are defined as the influencing factors of the local contrast, and the expression of the joint polarization enhancement coefficient is: ; in, is the degree of polarization in the entire window area, is the standard deviation of polarization in the entire window area.

[0039] Therefore, the polarization degree saliency map with the introduction of the joint polarization enhancement coefficient is: ; The entire image is traversed to obtain the polarization saliency map of the entire image.

[0040] Step 5: small target detection based on adaptive threshold.

[0041] Polarization saliency map can highlight dim targets, so adaptive threshold is used to segment dim targets from background. Defined as: ; in, and are the mean and variance of the polarization saliency map, is an adjustable constant parameter. Value greater than threshold When , the pixel is the target, otherwise it is the background.

[0042] Based on the above method, enhanced detection of small targets based on polarization consistency local contrast can be achieved. Small target detection in different scenarios can be achieved by continuously optimizing and adjusting parameters.

[0043] Based on the same inventive principle, this embodiment also provides a small target detection system based on polarization consistency local contrast, including a memory and one or more processors, wherein the memory stores executable code, and when one or more processors execute the executable code, it is used to implement the small target detection method mentioned in the above embodiment.

[0044] The embodiments described above provide a detailed description of the technical solutions and beneficial effects of the present invention. It should be understood that the above are only specific embodiments of the present invention and are not intended to limit the present invention. Any modifications, supplements and equivalent substitutions made within the scope of the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A small target detection method based on polarization consistency local contrast, characterized in that: The following steps are involved: (1) Obtain the grayscale image of the dim target and the corresponding multiple polarization images; (2) Based on the guided filtering algorithm, the grayscale image and polarization image are fused and denoised to obtain the output image; (3) Using the mechanism of local contrast and the joint polarization enhancement coefficient, the entire output image is traversed to obtain the polarization saliency map of the entire image; (4) In the polarization saliency map, an adaptive threshold is used to segment the dim target from the background to complete the dim target detection.

2. The method for detecting small targets based on polarization consistency local contrast according to claim 1, characterized in that: In step (1), the polarizer is rotated to change the angle of the analyzer. The rotation angles are 0°, 45°, 90°, and 135°, respectively. The corresponding four polarization images are: , , , .

3. The method for detecting small targets based on polarization consistency local contrast according to claim 1, characterized in that: The specific process of step (2) is as follows: The grayscale image is used as the guide image and the polarization image is used as the image to be filtered for guided filtering; for the non-target area, the gradient of the grayscale image is small, the output polarization image is the result of smoothing, and the noise is suppressed; for the target area, the gradient of the grayscale image is large, and the target is retained in the output polarization image.

4. The method for detecting small targets based on polarization consistency local contrast according to claim 1, characterized in that: The specific process of step (3) is as follows: Slide the sliding window from left to right and from top to bottom to traverse the entire image area and divide it into several sliding window areas; Each sliding window area is regarded as a sub-area, and each sub-area is divided into a central pixel area. and surrounding pixel areas; get the central pixel area Medium polarization maximum , mean polarization degree And the average polarization value of the surrounding pixel area , Indicates the number of the surrounding pixel area; Constructing joint polarization enhancement coefficient based on degree of polarization and polarization consistency ; Introducing the joint polarization enhancement factor The polarization degree significance map is: ； In the formula, Indicates the sequence number of the sliding window area, Indicates The polarization saliency map of the joint polarization enhancement coefficient is introduced in each sliding window region. Indicates the center pixel area and the Grayscale difference ratio of surrounding pixel areas; The entire image is traversed to obtain the polarization saliency map of the entire image.

5. The method for detecting small targets based on polarization consistency local contrast according to claim 1, characterized in that: In step (3), the joint polarization enhancement coefficient is constructed based on the polarization degree and polarization consistency , the formula is as follows: ； In the formula, is the polarization degree in the entire sliding window area; is the polarization standard deviation in the entire sliding window area, which represents the polarization consistency.

6. The method for detecting small targets based on polarization consistency local contrast according to claim 5, characterized in that: and The calculation formula is as follows: ； ； In the formula, Indicates light intensity information, and Represents linear polarization information, , and Obtained by the following formula: ； In the formula, Represents the value obtained by multiple polarization images Vector, Indicates light intensity information, and Represents linear polarization information, Represents circular polarization information. Since the ground-based polarization detection system does not collect circular polarization information, The weight is 0; , , , These are the four polarization images corresponding to the analysis angles of the rotating polarizer being 0°, 45°, 90°, and 135°.

7. The method for detecting small and dim targets based on polarization consistency local contrast according to claim 6, characterized in that: In step (5), the adaptive threshold Defined as: ； in, and are the mean and variance of the polarization saliency map, is an adjustable constant parameter. Value greater than threshold When , the pixel is the target, otherwise it is the background.

8. A small target detection system based on polarization consistency local contrast, characterized in that: The invention comprises a memory and one or more processors, wherein the memory stores executable codes, and when the one or more processors execute the executable codes, the method for detecting small and dim targets according to any one of claims 1 to 7 is implemented.

Images