Infrared small target detection method and system based on adaptive tradeoff structure tensor

Abstract

The invention provides an infrared small target detection method and system based on an adaptive tradeoff structure tensor, and relates to the technical field of infrared image target detection, and the method comprises the following steps: obtaining a to-be-processed infrared image containing a to-be-detected target, carrying out the preprocessing of the to-be-processed infrared image, and extracting a corresponding structure tensor; constructing an improved edge sensor and a corner sensing factor according to the structure tensor, and weighing the weight of the edge interference and the corner for the improved edge sensor and the corner sensing factor by adopting a structure similarity contrast quantity mode so as to better increase the identifiability of the target; a precise detection result can be obtained by carrying out simple threshold segmentation on the weighed result; according to the method, the target detection capability is remarkably improved while background clutters are suppressed, the significance of the target is kept, and the method has high detection rate and low false alarm rate.

Description

Technical Field

[0001] The present invention relates to the technical field of infrared image target detection, and more particularly to an infrared small target detection method and system based on adaptive trade-off structure tensor. Background Art

[0002] At present, infrared small target detection has been a research hotspot in recent years, and has broad application prospects in the fields of remote sensing, early warning devices and search systems. Compared with ordinary visible light, infrared imaging has more advantages in penetration. Compared with radar imaging, infrared imaging is more convenient and faster. However, its two characteristics of "weak" and "small" also make it a detection difficulty in research. "Weak" refers to the low signal-to-noise ratio of the target, poor contrast with the background, and weak infrared radiation intensity; and "small" refers to the small number of target pixels, making it difficult to obtain texture information during detection. The only information that can be considered is grayscale and position. The interference of background clutter is a major problem in detection. Therefore, the popularity of infrared small target detection under complex backgrounds has continued to rise in recent years.

[0003] However, the existing infrared small target detection methods are not stable enough when faced with complex backgrounds with a lot of clutter and clear high-contrast edges. The significant background edges also have sparse features that are difficult to suppress, and the threshold is usually set to a fixed value, which cannot flexibly separate the target from the complex background.

[0004] Therefore, how to provide an infrared small target detection method that can solve the above problems is an issue that those skilled in the art urgently need to solve. Summary of the invention

[0005] In view of this, the present invention provides an infrared small target detection method and system based on an adaptive trade-off structure tensor, which significantly improves the target detection capability while suppressing background clutter, maintains the target's significance, and has a high detection rate and a low false alarm rate.

[0006] In order to achieve the above object, the present invention adopts the following technical solution:

[0007] An infrared small target detection method based on adaptive trade-off structure tensor comprises the following steps:

[0008] Acquire an infrared image to be processed containing the target to be measured, and extract a corresponding structure tensor after preprocessing the infrared image to be processed;

[0009] Constructing an improved edge perceptron and a corner point perception factor according to the structure tensor, and constructing an adaptive edge suppression and target point enhancement model according to the improved edge perceptron and the corner point perception factor, and processing the infrared image to be processed using the adaptive edge suppression and target point enhancement model to simultaneously enhance the target according to the constrained background interference components that are adaptive to different scenes;

[0010] Using structural similarity to measure the weight of the background clutter edge component and the target component included in the processed infrared image to be processed;

[0011] The weighed results are segmented by threshold to obtain the detection results.

[0012] Preferably, it also includes:

[0013] An adjustable adaptive parameter based on the structural similarity metric is calculated, and the weights of the improved edge perceptron and the corner perception factor are balanced and controlled using the adjustable adaptive parameter.

[0014] Preferably, it also includes:

[0015] The adjustable adaptive parameters are designed to determine the best measurement method.

[0016] Preferably, the specific process of constructing the improved edge perceptron includes:

[0017] Acquire a first eigenvalue and a second eigenvalue of the infrared image to be processed through the structure tensor;

[0018] constructing an edge perception index according to the first eigenvalue and the second eigenvalue;

[0019] A nonlinear function is introduced to improve the edge perception index to obtain a final improved edge perceptron.

[0020] Preferably, the specific process of constructing the corner perception factor includes:

[0021] A corner perception factor is constructed according to the first eigenvalue and the second eigenvalue.

[0022] Preferably, the specific process of determining the optimal adjustable adaptive parameter includes:

[0023] Determining a value range and a step size of the adjustable adaptive parameter, and determining the number of the adjustable adaptive parameters to be tested according to the step size;

[0024] A structural similarity measurement index is constructed, and each adjustable adaptive parameter input value within a value range is calculated in the structural similarity measurement index, and the adjustable adaptive parameter with the smallest calculation result is selected as the optimal adjustable adaptive parameter.

[0025] The present invention also provides an infrared small target detection system based on an adaptive trade-off structure tensor, comprising:

[0026] An acquisition module is used to acquire an infrared image to be processed containing a target to be measured, and extract a corresponding structure tensor after preprocessing the infrared image to be processed;

[0027] A first construction module is used to construct an improved edge perceptron and a corner point perception factor according to the structure tensor, and to construct an adaptive edge suppression and target point enhancement model according to the improved edge perceptron and the corner point perception factor, and to process the infrared image to be processed using the adaptive edge suppression and target point enhancement model;

[0028] A second building module measures the weights of the background clutter edge components and the target components included in the infrared image to be processed after being processed by using the structural similarity measurement;

[0029] The detection module is used to perform threshold segmentation on the weighed results to obtain the detection results.

[0030] It can be seen from the above technical solutions that, compared with the prior art, the present invention discloses a method and system for infrared small target detection based on adaptive trade-off structure tensor, which has the following beneficial effects:

[0031] (1) An adaptive trade-off infrared small target detection method based on structure tensor is proposed. The structure tensor can be used alone to characterize the target and background edges, and small targets can be accurately detected with high efficiency in complex backgrounds.

[0032] (2) An improved edge perceiver (EO) is proposed, which can effectively target various clutter edges in the background and amplify the difference between them and the target. This enhancement is helpful for the subsequent separation process.

[0033] (3) An adaptive parameter α with adjustable important factors is set, which can adaptively select the value according to different types of background clutter, thereby enhancing the matching degree of the algorithm for processing edges, highlights and corner clutter.

[0034] The present invention was subsequently tested on six infrared sequences with different backgrounds. The results showed that the method provided by the present invention was always superior to other methods in SCRG and BSF scores, especially in dense cloud occlusion scenes. The experimental results show that the method provided by the present invention significantly improves the target detection capability while suppressing background clutter, maintains the saliency of the target, and has a high detection rate and a low false alarm rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0036] Figure 1 An overall flow chart of an infrared small target detection method based on an adaptive trade-off structure tensor provided by the present invention;

[0037] Figure 2 The embodiment of the present invention provides a comparison between the case where α is set to 0 and the case where α is set to an adaptive value, wherein the red frame is the target, the blue frame is the clutter, (a1)-(a2) are the original images, (b1)-(b2) show the target and clutter effects when α is set to 0, and (c1)-(c2) show the processing results of the adaptive value α;

[0038] Figure 3 A structural principle block diagram of an infrared small target detection system based on an adaptive trade-off structure tensor provided by the present invention;

[0039] Figure 4 A schematic diagram of the detection window and a schematic diagram of the background suppression factor results provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0040] 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.

[0041] See also Figure 1 As shown, an embodiment of the present invention discloses an infrared small target detection method based on an adaptive trade-off structure tensor, comprising the following steps:

[0042] Acquire an infrared image to be processed containing the target to be measured, and extract the corresponding structure tensor after preprocessing the infrared image to be processed, wherein the preprocessing process may be to perform Gaussian filtering on the infrared image to be processed;

[0043] An improved edge sensor and a corner point perception factor are constructed according to the structure tensor, an adaptive edge suppression and target point enhancement model is constructed according to the improved edge sensor and the corner point perception factor, and the adaptive edge suppression and target point enhancement model is used to process the infrared image to be processed, so as to enhance the target while adaptively constraining the background interference components according to different scenes;

[0044] The structural similarity is used to measure the weight of the background clutter edge component and the target component included in the processed infrared image to be processed;

[0045] The weighed results are segmented by threshold to obtain the detection results.

[0046] In a specific embodiment, it also includes:

[0047] An adjustable adaptive parameter based on a structural similarity metric is calculated, and the weights of the edge perception factor and the corner perception factor are improved by using the adjustable adaptive parameter balance control.

[0048] In a specific embodiment, the specific process of constructing an improved edge perceptron includes:

[0049] Acquire a first eigenvalue and a second eigenvalue of the infrared image to be processed through a structure tensor;

[0050] constructing an edge-aware index based on the first eigenvalue and the second eigenvalue;

[0051] The nonlinear function is introduced to improve the edge perception index and the final improved edge perceptron is obtained.

[0052] Specifically, the specific expression of the structure tensor is:

[0053]

[0054] In the formula, G β represents the Gaussian kernel, which is used to enhance the robustness of geometric direction estimation, and are the horizontal gradient and vertical gradient respectively; J β represents a symmetric positive semidefinite matrix, whose two eigenvalues ​​are λ1 and λ2 (i.e., the first eigenvalue λ1 and the second eigenvalue λ2), f σ Indicates that the infrared image to be processed is subjected to Gaussian smoothing filtering with a variance of σ, f σx 、f σy Respectively represent the gradient of Gaussian filtering in the x and y directions, J 11 , J 12 , J 21 , J 22 Represents the elements of a matrix.

[0055] In the calculation of the structure tensor, the first eigenvalue λ1 and the second eigenvalue λ2 contain the structural information of the local area centered on the pixel point. The first eigenvalue λ1 and the second eigenvalue λ2 show great feature differences in different areas of the infrared image. In flat areas, λ1≈λ2≈0, in edge areas such as clouds, λ1>>λ2≈0, and in corner areas, λ1≥λ2>>0.

[0056] Since the pixel value at the edge is much larger than that in the flat area and the corner area, an edge perception index (EI) is constructed. By defining the difference of eigenvalues, the grayscale value obtained in the edge area is much larger than that in the other two areas. The corresponding specific expression is:

[0057] EI(x,y)=λ1-λ2

[0058] Through experiments on edge perception indicators, it can be found that the background edges processed by the edge perception (EI) mostly show complete and continuous features, which is suitable for the situation where the target and background edges do not overlap.

[0059] However, one of the biggest challenges in detecting small infrared targets is to separate the complex background from the target. There is usually a lot of background clutter in a complex background. When a large number of clutter edges appear, it is somewhat redundant to use the edge perception index (EI) to process and highlight them all. In addition, these large numbers of edges will be filtered out when highlighting small targets in the next step. In this case, detailed and complete background edges increase the complexity of detection. For images with a large number of low-contrast edges and targets around the edges, the corner-like area of ​​the target is relatively weakened after processing with the edge perception index (EI), resulting in a decrease in the contrast between the target and the surrounding area, making it difficult to extract.

[0060] Therefore, in order to solve the above problem, the embodiment of the present invention introduces a nonlinear function e x The edge perception index is improved to obtain the final improved edge perceptron. The specific expression is:

[0061]

[0062] Nonlinear function e x has an important property, that is, when the independent variable x increases, e x The growth rate of will also accelerate, which means that for larger λ1-λ2 values ​​(i.e., high-contrast edges), the improved edge perceptron will give a larger value, thereby emphasizing these edges more strongly. The growth law presented by the nonlinear function is "explosive growth", so for smaller λ1-λ2 values ​​(i.e., low-contrast edges), the grayscale value will not be greatly enhanced. The low-contrast edges processed by this formula can be weakened on the image, and the final image retains the target and the strong edges that are difficult to remove with high contrast, which reduces the workload for subsequent data processing, and the low-contrast edges around the target in the processed images of some scenes will also be reduced.

[0063] In a specific embodiment, the specific process of constructing the corner perception factor includes:

[0064] A corner perception factor is constructed according to the first eigenvalue λ1 and the second eigenvalue λ2.

[0065] Specifically, in the detection of weak infrared targets, the edge area and the corner area in the image coexist, and this feature is also reflected in the target, and the target area also contains the features of the edge and the corner area. Using only the improved edge sensor processing can only achieve good processing for images with clear background edges, but cannot achieve the recognition of weak targets when there are a large number of corner points and the edge is blurred and the contrast is not strong in the image. Therefore, the embodiment of the present invention constructs a corner point perception indicator, and the corresponding specific expression is:

[0066]

[0067] In the embodiment of the present invention, the corner perception index constructed using the concept of the harmonic mean can suppress the high brightness noise point (PNHB) type outliers appearing in the background by taking the reciprocal form, and can highlight targets with low contrast.

[0068] The target area after processing with the improved edge sensor loses the characteristics of the corner area. When the target occupies a large number of pixel values ​​in the image, there will be only one outer edge of the target. When the target occupies a small proportion of pixel values ​​in the image, the target will be lost, thereby reducing the detection rate. Different proportions of edge and corner perception should be used for infrared weak target images in different scenes. For the case where the background clutter is relatively flat and the target occupies a small proportion of pixel values ​​in the image, using the corner perception indicator for processing can obtain a higher detection rate and faster detection speed.

[0069] In a specific embodiment, it also includes:

[0070] The adjustable adaptive parameter α is designed to determine the best measurement method.

[0071] In a specific embodiment, the specific process of determining the optimal adjustable adaptive parameter α includes:

[0072] Determine the value range and step size of the adjustable adaptive parameter, and determine the number of adjustable adaptive parameters to be tested according to the step size;

[0073] A structural similarity measurement index is constructed, and each adjustable adaptive parameter input value within a value range is calculated in the structural similarity measurement index, and the adjustable adaptive parameter with the smallest calculation result is selected as the optimal adjustable adaptive parameter.

[0074] In a specific embodiment, the specific process of determining the optimal adjustable adaptive parameter includes:

[0075] Determine the value range and step size of the adjustable adaptive parameter, and determine the number of adjustable adaptive parameters to be tested according to the step size;

[0076] A structural similarity measurement index is constructed, and each adjustable adaptive parameter input value within a value range is calculated in the structural similarity measurement index, and the adjustable adaptive parameter with the smallest calculation result is selected as the optimal adjustable adaptive parameter.

[0077] Specifically, the embodiment of the present invention proposes an adjustable adaptive parameter α, which is used to control the weights of the improved edge perception sensor and the corner perception indicator. The corresponding specific expression is:

[0078] ATST(x,y)=α*EO+(1-α)*CI

[0079] In the formula, CI represents the corner perception indicator, EO represents the improved edge perception, α is the adaptive weight value, and the value range is [0,1]. The implementation process is shown in Table 1:

[0080] Table 1 Implementation process of α

[0081] Algorithm 1: Calculate adaptive weight value Input: original image Output: ATST type <![CDATA[1. Calculate the image eigenvalue λ1 and λ2 according to formulas (3) and (4)]]> 2. Calculate the improved edge perceptron numerical EO according to formula (4) 3. According to formula (5), calculate the corner perception indicator value CI 4. Input the value obtained in 1-3 into formula (6) to obtain the ATST formula of the unknown coefficient α 5. Input the ATST formula of the unknown coefficient α into Algorithm 2

[0082] The basis for selecting the value of the adaptive parameter α is to traverse 100 values ​​in the range of 0-1 and select the α value with the best processing effect. The best effect is to substitute the current α value into the structural similarity index measure (SSIM) method proposed in the embodiment of the present invention for measurement, and judge which indicator and the required weight are more suitable for the image according to the processing result, and finally determine the best α value suitable for the current image.

[0083] The structural similarity index measure (SSIM) is usually expressed as three factors: image correlation loss, brightness distortion, and contrast distortion. The specific expression is:

[0084] SSIM(f,g)=l(f,g)c(f,g)s(f,g)

[0085]

[0086] In the formula, the first term of l(f,g) is the brightness comparison function, which is used to measure the average brightness of the two images (μ f and μ g ), this factor is only f =μ gThe second term c(f,g) is the contrast comparison function, which is used to measure the closeness of the contrast between two images. The contrast is based on the standard deviation σ f and σ g Only when σ f =σ g When , this term is maximum and equal to 1. The third term s(f,g) is the structural comparison function, which is used to represent the correlation coefficient between two images f and g, where σ fg is the covariance between f and g, the positive value of the SSIM index is in [0,1], 0 means there is no correlation between the images, 1 means f = g, C1, C2 and C3 are constants.

[0087] Using structural similarity as a measurement indicator, the weight value when the current image has the smallest structural similarity with the original image is the value we need. The evaluation process is shown in Table 2:

[0088] Table 2 Specific implementation process of evaluating α

[0089]

[0090] The embodiment of the present invention designs two cases where α is 0 and the case where α is adaptively selected according to the algorithm for comparison, and the images in the data set e are used. Figure 2 From (b1)-(b2), we can see that when the weight α is 0, the background clutter (blue box area) in the image still exists in large quantities. When observing its three-dimensional image, the clutter part is still significant. Figure 2 From (c1)-(c2), we can see that the background clutter is well removed. Although the range of the target is also reduced compared to (b1)-(b2), we can see from (c1)-(c2) that the saliency of the target has not weakened, and it can still be distinguished from the background and accurately identified, and the clutter can be removed.

[0091] See also Figure 3 As shown, the present invention also provides an infrared small target detection system based on an adaptive trade-off structure tensor, comprising:

[0092] An acquisition module is used to acquire an infrared image to be processed containing a target to be measured, and extract a corresponding structure tensor after preprocessing the infrared image to be processed;

[0093] A first construction module is used to construct an improved edge perceptron and a corner point perception factor according to the structure tensor, and to construct an adaptive edge suppression and target point enhancement model according to the improved edge perceptron and the corner point perception factor, and to process the infrared image to be processed using the adaptive edge suppression and target point enhancement model;

[0094] The second building module measures the weight of the background clutter edge component and the target component included in the processed infrared image to be processed by using the structural similarity measurement;

[0095] The detection module is used to perform threshold segmentation on the weighed results to obtain the detection results.

[0096] In order to evaluate the performance of the method provided by the embodiment of the present invention, the embodiment of the present invention uses four evaluation indicators widely used in the field of infrared small target detection: PD, FA, SCR, BSF. Among them, the most important indicators for evaluating the success rate of target detection are PD and FA, which are defined as follows:

[0097]

[0098] The curve drawn with Fa as the horizontal coordinate and Pd as the vertical coordinate is called the ROC curve. The area under the curve (AUC) can quantify the detection performance. The larger the area under the ROC curve, the better the detection performance. The detection rate (Pd) reflects the detection ability of the algorithm, and the false alarm rate (Fa) indicates the false detection rate of the algorithm.

[0099] The signal-to-clutter ratio (SCR) can reflect the difficulty of detecting small targets, and the signal-to-clutter ratio gain (SCRG) is used to evaluate the ability of the algorithm to enhance the signal and describe the significance of the target enhancement after the algorithm is processed. The formula used is as follows:

[0100]

[0101] In the formula, μ t is the average pixel value of the target, μ b and σ b are the average pixel value and standard deviation of the pixel values ​​in the neighboring area around the target. When the detection meets the following two conditions, it means that the detected target is a real target: 1) The detection result overlaps with the real target in the image. 2) The central pixel position of the detection result and the central pixel position of the real target do not differ by more than a certain range, that is, 5 pixels. The schematic diagram of the method used for detection in the target area is shown in Figure 4 shown.

[0102] The background suppression factor (BSF) is used to indicate the suppression ability of the background after algorithm processing, and its definition is as follows:

[0103]

[0104] In the formula, σ in and σ outThey represent the background standard deviation of the input image and the standard deviation of the output image background respectively. The size of BSF reflects the algorithm's ability to suppress the background. The larger the value, the stronger the suppression ability. Similarly, the larger the value of SCRG, the more effective the algorithm is in enhancing the target.

[0105] The embodiment of the present invention uses 6 infrared small target detection sequences under different backgrounds for comparative evaluation. Sequence a has a dark background with low contrast, and the small target is hidden in dense grass and has partial overlap; sequence b has many striped clouds in the background, and some small targets are hidden in the clouds; sequence c has a high background contrast, and the clouds are dense and fluctuating violently; sequence d has an open background, but the contrast between the small target and the surrounding area is very low, and there is a large dark area in the background; sequence e has clouds with similar grayscale to the small target, which is prone to false detection; sequence f has a lot of noise in the background, and the noise intensity is close to that of the small target. The specific experimental conditions are shown in Table 3

[0106] Table 3 Infrared sequence characteristics

[0107]

[0108]

[0109] The embodiment of the present invention evaluates the effectiveness of the algorithm and compares it with four different baseline methods. At the same time, four different types of infrared small target detection algorithms are selected for comparative tests, namely LR, LEF, MPCM, and TLLCM. The specific results are shown in Tables 4 and 5. It can be found from Table 4 that the method proposed in the embodiment of the present invention has the highest SCRG score in five scenes, and far exceeds the scores of other baseline methods in some scenes. The small target in sequence (c) has low brightness and poor imaging. The method provided by the embodiment of the present invention scores slightly lower than the local contrast method. However, in a complex occluded background, the method proposed in the embodiment of the present invention can fully improve the intensity of small targets. It can be seen from Table 5 that the method of the embodiment of the present invention is comparable to most algorithms in some scenes, and far exceeds the data of other algorithms in half of the scenes.

[0110] Table 4 Comparison of SCRG scores of different methods

[0111]

[0112] Table 5. Comparison of BSF scores of different methods. The best scores are marked in bold.

[0113]

[0114] In summary, the method proposed in the embodiment of the present invention can be adapted to most scenes, especially in scenes blocked by dense clouds, where the SCRG and BSF scores are higher than those of the comparison algorithm, and can suppress complex backgrounds while enhancing target saliency.

[0115] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0116] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for detecting small infrared targets based on adaptive trade-off structure tensor, characterized in that: The following steps are involved: Acquire an infrared image to be processed containing the target to be measured, and extract a corresponding structure tensor after preprocessing the infrared image to be processed; constructing an improved edge perceptron and a corner point perception factor according to the structure tensor, constructing an adaptive edge suppression and target point enhancement model according to the improved edge perceptron and the corner point perception factor, and processing the infrared image to be processed using the adaptive edge suppression and target point enhancement model; Using structural similarity to measure the weight of the background clutter edge component and the target component included in the processed infrared image to be processed; The weighed results are segmented by threshold to obtain the detection results.

2. The infrared small target detection method based on adaptive trade-off structure tensor according to claim 1 is characterized in that: Also includes: An adjustable adaptive parameter based on the structural similarity metric is calculated, and the weights of the improved edge perceptron and the corner perception factor are balanced and controlled using the adjustable adaptive parameter.

3. The infrared small target detection method based on adaptive trade-off structure tensor according to claim 2 is characterized in that: Also includes: The adjustable adaptive parameters are designed to determine the best measurement method.

4. The infrared small target detection method based on adaptive trade-off structure tensor according to claim 2 is characterized in that: The specific process of building an improved edge perceptron includes: Acquire a first eigenvalue and a second eigenvalue of the infrared image to be processed through the structure tensor; constructing an edge perception index according to the first eigenvalue and the second eigenvalue; A nonlinear function is introduced to improve the edge perception index to obtain a final improved edge perceptron.

5. The infrared small target detection method based on adaptive trade-off structure tensor according to claim 4 is characterized in that: The specific process of constructing the corner perception factor includes: A corner perception factor is constructed according to the first eigenvalue and the second eigenvalue.

6. The infrared small target detection method based on adaptive trade-off structure tensor according to claim 3 is characterized in that: The specific process of determining the best adjustable adaptive parameters includes: Determining a value range and a step size of the adjustable adaptive parameter, and determining the number of the adjustable adaptive parameters to be tested according to the step size; A structural similarity measurement index is constructed, and each adjustable adaptive parameter input value within a value range is calculated in the structural similarity measurement index, and the adjustable adaptive parameter with the smallest calculation result is selected as the optimal adjustable adaptive parameter.

7. A system using the infrared small target detection method based on adaptive trade-off structure tensor according to any one of claims 1 to 6, characterized in that: include: An acquisition module is used to acquire an infrared image to be processed containing a target to be measured, and extract a corresponding structure tensor after preprocessing the infrared image to be processed; A first construction module is used to construct an improved edge perceptron and a corner point perception factor according to the structure tensor, and to construct an adaptive edge suppression and target point enhancement model according to the improved edge perceptron and the corner point perception factor, and to process the infrared image to be processed using the adaptive edge suppression and target point enhancement model; A second building module measures the weights of the background clutter edge components and the target components included in the infrared image to be processed after being processed by using the structural similarity measurement; The detection module is used to perform threshold segmentation on the weighed results to obtain the detection results.

Images