Infrared camouflage performance evaluation method and device

Abstract

The invention relates to the technical field of infrared performance evaluation, in particular to an infrared camouflage performance evaluation method and device. The method comprises the steps that an infrared image is acquired, and the infrared image carries a target and a background; feature extraction is carried out on the infrared image, a feature set is generated, and pixel-level features corresponding to the infrared image and internal texture structure features corresponding to the target are carried in the feature set; determining the main saliency of the target based on the pixel-level features corresponding to the infrared image; determining the internal saliency of the target based on the internal texture structure features corresponding to the target; and determining an infrared camouflage efficiency value of the target based on the main saliency and the internal saliency. By applying the method, the infrared camouflage efficiency of the target can be measured and analyzed.

Description

Technical Field

[0001] The present application relates to the field of infrared effectiveness evaluation technology, and in particular to an infrared camouflage effectiveness evaluation method and device. Background Art

[0002] Infrared camouflage effectiveness refers to the effect of reducing the detection and identification ability of infrared sensors after the target takes camouflage measures in the infrared band. The target can be any physical object such as a car.

[0003] However, existing methods mainly quantify and evaluate the differences in radiation characteristics between the target and the background, and are unable to analyze and evaluate from dimensions such as target image features, target feature fusion and saliency, and target internal saliency.

[0004] In view of this, how to provide an infrared camouflage effectiveness evaluation method to achieve quantitative measurement and analysis of the target's infrared camouflage effectiveness has become a technical problem that urgently needs to be solved. Summary of the invention

[0005] The embodiments of the present application provide an infrared camouflage effectiveness evaluation method, an infrared camouflage effectiveness evaluation device, an electronic device and a computer storage medium, which are used to solve the current problem that it is difficult to quantify and analyze the infrared camouflage effectiveness of a target.

[0006] In a first aspect of an embodiment of the present application, a method for evaluating infrared camouflage effectiveness is provided, comprising:

[0007] Acquire an infrared image, wherein the infrared image carries a target and a background;

[0008] Extracting features from the infrared image to generate a feature set, wherein the feature set carries pixel-level features corresponding to the infrared image and internal texture structure features corresponding to the target;

[0009] Determining the primary saliency of the target based on the pixel-level features corresponding to the infrared image;

[0010] Determining the internal saliency of the target based on the internal texture structure features corresponding to the target;

[0011] Based on the main saliency and the internal saliency, an infrared camouflage effectiveness value of the target is determined.

[0012] In a second aspect of an embodiment of the present application, a method for evaluating infrared camouflage effectiveness is provided, comprising:

[0013] An acquisition module is configured to acquire an infrared image, wherein the infrared image carries a target and a background;

[0014] An extraction module is configured to perform feature extraction on the infrared image to generate a feature set, wherein the feature set carries pixel-level features corresponding to the infrared image and internal texture structure features corresponding to the target;

[0015] A first determination module is configured to determine the primary saliency of the target based on the pixel-level features corresponding to the infrared image;

[0016] A second determination module is configured to determine the internal saliency of the target based on the internal texture structure features corresponding to the target;

[0017] The third determination module is configured to determine the infrared camouflage effectiveness value of the target based on the main saliency and the internal saliency.

[0018] According to a third aspect of an embodiment of the present application, a computing device is provided, including:

[0019] Memory and processor;

[0020] The memory is used to store computer executable instructions, and the processor is used to execute the computer executable instructions. When the computer executable instructions are executed by the processor, the steps of the above-mentioned infrared camouflage effectiveness evaluation method are implemented.

[0021] According to a fourth aspect of an embodiment of the present application, a computer-readable storage medium is provided, which stores computer-executable instructions, and when the instructions are executed by a processor, the steps of the above-mentioned infrared camouflage effectiveness evaluation method are implemented.

[0022] The present application provides an infrared camouflage effectiveness evaluation method, comprising: first acquiring an infrared image, wherein the infrared image carries a target and a background; then performing feature extraction on the infrared image to generate a feature set, wherein the feature set carries pixel-level features corresponding to the infrared image and internal texture structure features corresponding to the target; secondly determining the primary saliency of the target based on the pixel-level features corresponding to the infrared image; thirdly determining the internal saliency of the target based on the internal texture structure features corresponding to the target; and finally determining the infrared camouflage effectiveness value of the target based on the primary saliency and the internal saliency.

[0023] By applying the infrared camouflage effectiveness evaluation method provided in the embodiment of the present application, the infrared camouflage effectiveness of the target is reflected through the final determined infrared device effectiveness value, thereby achieving quantitative measurement and analysis of the infrared camouflage effectiveness of the target.

[0024] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, which can be implemented in accordance with the contents of the specification, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present application. Also, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings:

[0026] Figure 1 A flow chart of an infrared camouflage effectiveness evaluation method provided in an embodiment of the present application;

[0027] Figure 2 A schematic diagram of brightness characteristics in an infrared camouflage effectiveness evaluation method provided in an embodiment of the present application;

[0028] Figure 3 A schematic diagram of gradient features in an infrared camouflage effectiveness evaluation method provided in an embodiment of the present application;

[0029] Figure 4 A schematic diagram of contrast characteristics in an infrared camouflage effectiveness evaluation method provided in an embodiment of the present application;

[0030] Figure 5 A schematic diagram of the calculation process of the LBP feature in an infrared camouflage effectiveness evaluation method provided in an embodiment of the present application;

[0031] Figure 6 A schematic diagram of LBP characteristics in an infrared camouflage effectiveness evaluation method provided in an embodiment of the present application;

[0032] Figure 7 A schematic diagram of entropy characteristics in an infrared camouflage effectiveness evaluation method provided in an embodiment of the present application;

[0033] Figure 8 A schematic diagram of orientation characteristics in an infrared camouflage effectiveness evaluation method provided in an embodiment of the present application;

[0034] Fig. 9 A schematic diagram of angular density characteristics in an infrared camouflage effectiveness evaluation method provided in an embodiment of the present application;

[0035] Fig.10 A schematic diagram of the structure of an infrared camouflage effectiveness evaluation device provided in an embodiment of the present application;

[0036] Fig.11A structural block diagram of a computing device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0037] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0038] See also Figure 1 , Figure 1 A schematic diagram of a flow chart of an infrared camouflage effectiveness evaluation method provided in an embodiment of the present application. Figure 1 As shown, the specific steps include the following steps.

[0039] Step S102: Acquire an infrared image, wherein the infrared image carries a target and a background.

[0040] Step S104: extracting features from the infrared image to generate a feature set, wherein the feature set carries pixel-level features corresponding to the infrared image and internal texture structure features corresponding to the target.

[0041] It should be noted that the pixel-level features here include: brightness features, gradient features, contrast features, LBP features, entropy features, orientation features and corner density features.

[0042] In an embodiment of the present application, the feature extraction of the infrared image and the generation of a feature set include: establishing an infrared camouflage effectiveness evaluation index system for the target, wherein the infrared camouflage effectiveness evaluation index system is used to select image features for evaluating the infrared camouflage effectiveness of the target.

[0043] After extracting features from the infrared image and generating a feature set, the method further includes: converting high-dimensional feature information in the feature set into low-dimensional feature information.

[0044] In practical applications, before extracting features from infrared images, it is necessary to first establish an infrared camouflage effectiveness evaluation index system for the target, that is, to determine which image features to select for the infrared camouflage effectiveness evaluation of the target.

[0045] The image features of a target refer to the information contained in the image that can be used to describe one or more properties of the target. In order to reduce the dimensionality of image feature information, appropriate mathematical methods (mapping or transformation methods) are also needed to convert high-dimensional feature information into low-dimensional feature information.

[0046] According to the human eye visual attention mechanism, in visible light images, image features such as image gradient, texture and color can well reflect the structural information of the image and are important parameter indicators for calculating image saliency; when Harel proposed the GBVS (Graph-Based Visual Saliency) algorithm, he selected three types of features, namely color, brightness and texture orientation, to calculate the saliency map of visible light images. However, infrared images are grayscale images, so color features are not desirable. In addition, infrared images have lower imaging accuracy than visible light images, and their details such as edges, shapes and textures are not clear, and the intuitive image features that can be directly used are even more insufficient. In order to more effectively apply the GBVS algorithm to the infrared camouflage effectiveness analysis of the target, the present invention makes adaptive improvements to the algorithm. When designing the feature vector, on the one hand, "pixel-level features" are selected, and on the other hand, the internal texture structure features (i.e., internal shape features) of the target are selected; these two major feature categories are combined to form the infrared camouflage effectiveness evaluation index system of the target.

[0047] The pixel-level features refer to those image features that can be converted into a numerical value at a pixel position. In the present invention, the pixel-level features include: brightness features, gradient features, contrast features, LBP features, entropy features, orientation features, and corner density features.

[0048] As for shape features, in real life, in the human visual perception system, shape (including various triangles, rectangles, parallelograms and even complex polygonal structures) is a concept that is difficult to accurately define but widely understood. In a computer system, in order to characterize the shape information of an object, either the edge contour of the object is outlined by the contour of feature points, or the overall shape features of the object are described based on the area occupied by the object in the image. In the present invention, the line segment detection result inside the target in the infrared image is used as a consideration for the target camouflage effectiveness analysis, and based on the experimental results of predecessors, the concept of linearity is introduced into the target's infrared camouflage effectiveness evaluation index system.

[0049] In the process of extracting features from infrared images and obtaining feature extraction results, the extraction mechanisms of these pixel-level features, such as brightness features, gradient features, contrast features, LBP features, entropy features, orientation features, and corner density features, are as follows:

[0050] (1) Brightness characteristics

[0051] Because the infrared image is a grayscale image, the grayscale value of each pixel is the brightness feature I of the pixel.

[0052] In a specific example, the calculated infrared image brightness feature map is shown in Figure 2 , Figure 2 A schematic diagram of brightness characteristics in an infrared camouflage effectiveness evaluation method provided in an embodiment of the present application.

[0053] like Figure 2 As shown, T1, T2 and T3 are three targets respectively, the four images in the first column correspond to T1, the four images in the middle column correspond to T2, and the four images on the right correspond to T3.

[0054] (2) Gradient features

[0055] Because the image is a discrete digital signal, the difference is used to replace the differential in the continuous signal. The gradient itself is a vector with both magnitude and direction. In the present invention, the modulus of the gradient is used. The modulus |M(i, j)| of the gradient corresponding to the pixel P(i, j) in the image is:

[0056]

[0057] In a specific example, the calculated gradient feature map is shown in Figure 3 , Figure 3 This is a schematic diagram of gradient features in a method for evaluating infrared camouflage effectiveness provided in an embodiment of the present application. It should be noted that: Figure 2 and Figure 3 correspond to each other, so in Figure 3 The four images in the first column also correspond to T1, the four images in the middle column also correspond to T2, and the four images on the right also correspond to T3.

[0058] (3) Contrast characteristics

[0059] δ(i, j) = |ij| represents the grayscale difference between adjacent pixels, P δ (i, j) represents the probability distribution of pixels with grayscale difference of adjacent pixels being δ, and the calculation formula of contrast C is:

[0060]

[0061] In order to extract the contrast features of each layer of the pyramid image, it is necessary to apply the local contrast information of the central pixel. In actual operation, a fixed-size pixel template is often used to traverse the image, and then the difference between the grayscale value of the central pixel of the template and the grayscale value of the surrounding neighborhood pixels is calculated in sequence. The sum of the squares of these differences is calculated and then divided by the number of the above square terms, and the local contrast information of each pixel in each layer of the image can be obtained.

[0062] In a specific example, the calculated contrast feature map is shown in Figure 4 , Figure 4 A schematic diagram of contrast characteristics in an infrared camouflage effectiveness evaluation method provided in an embodiment of the present application.

[0063] (4) LBP features

[0064] The original LBP operator is defined as follows: a 3×3 template is defined in the original image, and the central pixel of the template window is used as the threshold Thr. The grayscale values ​​of the 8 pixels adjacent to the central pixel are compared with Thr. If the surrounding pixel value is greater than Thr, the position of the pixel is marked as 1, otherwise it is marked as 0. Through such calculation, the 8 surrounding points in the 3×3 neighborhood of the central pixel of the template can be compared with Thr to generate an 8-bit binary number (for convenience, it can also be converted into a decimal number, namely LBP code, a total of 256 types), so that the LBP value of the central pixel of the window is obtained. If expressed in mathematical language, the calculation formula of the LBP code is:

[0065] Among them, g(a j -a i ) is a function defined as follows:

[0066] The above scheme is explained below with examples. In a specific example, when the LBP code of a certain pixel in the infrared image is equal to 01001100, the corresponding decimal LBP code value is 76; see the calculation process for a schematic diagram. Figure 5 , Figure 5 The diagram of the calculation process of the LBP feature in the infrared camouflage effectiveness evaluation method provided in the embodiment of the present application is shown in the figure. The leftmost table is a template of size 3×3, so the threshold Thr is equal to 6. The 4 in the template is compared with the threshold Thr(6). Since 4 is less than 6, the corresponding position in the left table (i.e., the first row and the first column) is marked as 0, and the same is true for other pixels. In this way, the table on the right is obtained, and the values ​​in the table constitute an 8-bit binary number (i.e., 01001100). Finally, the decimal LBP code corresponding to 01001100 (i.e., 76) is calculated.

[0067] In a specific example, the calculated LBP feature map is shown in Figure 6 , Figure 6 A schematic diagram of LBP characteristics in an infrared camouflage effectiveness evaluation method provided in an embodiment of the present application.

[0068] (5) Entropy characteristics

[0069] According to Shannon's formula, the entropy H(X) of a discrete random variable X is:

[0070] The entropy of an image represents the average amount of information in the image, which is its physical meaning. To expand on this, the one-dimensional entropy of an image specifically represents the amount of information contained in the aggregated features of the grayscale distribution in the image. If the grayscale values ​​of each pixel in the image are statistically classified, and the number and probability of each grayscale value are calculated, the one-dimensional grayscale entropy of the image can be obtained based on this:

[0071] However, it is obvious that the one-dimensional entropy of an image cannot reflect the spatial characteristic information of the grayscale distribution of the image. At this point, it becomes necessary to introduce a feature quantity that can reflect the spatial characteristics of the grayscale distribution - the two-dimensional entropy of the image. The key to calculating the two-dimensional entropy of an image is to construct a feature binary (i, j), where i represents the pixel grayscale of the target pixel itself, and j represents the grayscale mean of the neighborhood of the target pixel in the image.

[0072] In infrared images, i represents the gray value of a pixel (0<=i<=255), j represents the neighborhood gray mean (0<=j<=255), then the comprehensive characteristics of the gray value at the target pixel position and the gray distribution of its neighborhood pixels can be constructed as follows: ij =f(i,j) / N 2 .

[0073] Among them, f(i,j) is the frequency of occurrence of feature binary (i,j), and N is the scale of the image.

[0074] Then the discrete two-dimensional entropy of the image can be defined as:

[0075] From a physical perspective, the two-dimensional entropy of an image can not only represent the amount of information contained in the image, but also highlight the comprehensive characteristics of the grayscale information of the pixel position in the image and the grayscale distribution in the pixel neighborhood. It is a very useful image feature.

[0076] In a specific example, the entropy feature map calculated is shown in Figure 7 , Figure 7 A schematic diagram of entropy characteristics in an infrared camouflage effectiveness evaluation method provided in an embodiment of the present application.

[0077] (6) Orientation characteristics

[0078] It can also be called Gabor texture feature, which is a mathematical method commonly used for texture orientation analysis.

[0079] In the field of human vision, after a large number of physiological and psychological experiments, scientists have discovered that when the human eye observes an image, the first thing that can be directly observed is the low-level visual features in the image, but when the human brain processes these image signals, it actually decomposes the signals into a series of channels with different frequencies and orientations through linear spatial filters, so that the image perception effect can be formed in the human brain, completing the mapping of primary visual signals to brain signals. This process completes the decomposition of image signals, and based on this process, new image features that are different from low-level visual features can be derived, so it is very important to mathematically model this process.

[0080] In the actual operation process, people have proposed a method to extract the orientation features of images based on Gabor transform. The reason why Gabor transform shines in the field of image processing is due to its two important advantages: one is its good localization properties in the spatial domain and frequency domain; the other is that no matter from the fluctuation characteristics of the spatial domain, the orientation selection characteristics, the spatial domain and frequency domain selection, or the orthogonal phase relationship, the two-dimensional Gabor basis function can effectively simulate the visual stimulus response of simple cells in the visual epidermis of most mammals.

[0081] In the present invention, after acquiring the infrared image, based on multiple Gabor filter groups, the infrared image is filtered at different center frequencies and orientations to extract image features containing different frequency components and different orientation information. Here, in order to speed up the calculation efficiency, several center frequencies and orientations are usually selected based on experience in the actual operation process as parameters of the Gabor filter.

[0082] The spatial domain formula of the one-dimensional Gabor filter is:

[0083]

[0084] Similar to the one-dimensional Gabor filter, the spatial domain of the two-dimensional Gabor filter is described as:

[0085]

[0086] Among them, σ x and σ y Represents the spatial scale factors for horizontal and vertical directions, ω f and θ f Represent the center frequency and azimuth respectively.

[0087] According to research results in the field of biology, human nerve cells have a certain response frequency in response to a certain physical target in the spatial field. Scientists call this range of response frequency frequency bandwidth, and can use half-peak bandwidth to characterize this phenomenon (referring to the width value obtained by half of the maximum response amplitude value of human visual nerve cells). According to calculations, the response range of human visual nerve cells is generally between 0.6 octaves and 1.3 octaves, and the response mean is generally 1.3 octaves. After a lot of practical analysis, previous experimental results show that when the half-peak bandwidth is 1, the filter presents a stable and good effect. Assuming the given half-peak bandwidth is, the center frequency is, the mutual constraint relationship between the two and the horizontal change scale is as follows:

[0088]

[0089] The spatial aspect ratio of the Gabor filter can be expressed as: λ = σ x / σ y .

[0090] Experience shows that the smaller the λ value, the stronger the directional characteristics of the filter. Under normal circumstances, its optimal value range is [0.23, 0.92]. As for the center frequency range of the Gabor filter, the upper and lower limits are defined by:

[0091]

[0092] Where i = 1, 2, ..., log2(N c / 8); N c is the width of the input image.

[0093] In practical applications, the orientation parameter θ∈{0°, 45°, 90°, 135°} can determine the Gabor feature information of the target in the θ orientation. At this point, the parameters of a Gabor filter bank are all determined.

[0094] In a specific example, the calculated orientation feature map is shown in Figure 8 , Figure 8 This is a schematic diagram of orientation characteristics in a method for evaluating the effectiveness of infrared camouflage provided in an embodiment of the present application. It should be noted that: Figure 2 When the orientation feature parameters are 0°, 45°, 90° and 135° respectively, the orientation feature extraction results of the infrared images where T1, T2 and T3 are located are as follows: Figure 8 shown.

[0095] (7) Corner density feature

[0096] Corner points refer to points with particularly prominent attributes in certain aspects (such as extreme points), and are often used in the field of mathematics. For images, corner points are the pixels corresponding to the local maximum of the first-order derivative (the gradient of grayscale images). In images, the shape of corner points is often manifested as the intersection of two or more edges. From a mathematical point of view, it can be considered that the position near the corner point in the image is the area where the gradient value and gradient direction change very quickly. The physical meaning of corner points is: on the one hand, people believe that the superiority of corner points lies in the ability to effectively reduce the amount of information data while retaining image feature information, which is of great significance to solving the reliable matching of images; on the other hand, for the same scene, even if the viewing angle changes, the corner points are usually unchanged and have good stability.

[0097] In order to better explore the target feature information in the infrared image, the present invention introduces the concept of corner density feature. Corner detection is performed on an infrared image, and then the image is divided into many super-pixel blocks. The number of corner points in each pixel block is counted, and then divided by the size of the super-pixel block, the corner density feature in a certain pixel block can be obtained. This means that the mathematical definition of the so-called corner density feature is as follows:

[0098]

[0099] Where D(P) represents the corner density within the superpixel block P, NUM c represents the number of corner points in the superpixel block P, w and h represent the width and height of the superpixel block P respectively.

[0100] In order to effectively extract the corner density features of the target in the infrared image, the Harris corner detection algorithm is used to detect the corners in the infrared image. The core idea of ​​the algorithm is to move a small window in the image and determine the position of the corner by examining the average transformation value of the image grayscale in the small window; the specific mathematical description is as follows:

[0101] Assume that the image window translation amount is (x, y) and the resulting grayscale change is E(x, y), then:

[0102]

[0103] Among them, w u,v It is a window function, which is often set to the form of a Gaussian function in practical applications and plays a weighting role; I x+u,y+u is the gray value after translation; I u,v is the gray value before translation. Taylor expansion of the above formula gives:

[0104]

[0105] Among them, A, B, and C are approximations of second-order directional differentials, which can be expressed as:

[0106]

[0107] Among them, G(x,y) represents the specific form of the Gaussian smoothing filter function used in the actual operation; X, Y represent the first-order directional differential. Specifically, the x-differential operator is |1 0 -1|, and the y-differential operator is |1 0 -1 T , then the specific expressions of X and Y can be written as:

[0108]

[0109] According to the relevant knowledge of matrix theory, E(x, y) can be further expressed as follows:

[0110]

[0111] Among them, the matrix M is the approximate Hessian matrix of the autocorrelation function E(x,y):

[0112]

[0113] At this point, the eigenvalue of the matrix M can be considered to approximately represent the extreme curvature of the grayscale autocorrelation function at a certain point in the image. Then, it means that if the two eigenvalues ​​of the matrix M are relatively large, it means that the curvature extremes in two orthogonal directions of the image grayscale autocorrelation function at the pixel point are both large, and the current pixel point can be considered as a corner point.

[0114] In a specific example, the calculated corner density feature map can be found in Fig. 9 , Fig. 9 This is a schematic diagram of angular density characteristics in an infrared camouflage effectiveness evaluation method provided in an embodiment of the present application. It should be noted that: Figures 2 to 9 are mutually corresponding, that is, Figures 2 to 9 The processing objects are the same.

[0115] Step S106: determining the primary saliency of the target based on the pixel-level features corresponding to the infrared image.

[0116] In the embodiment of the present application, the primary saliency of the target is determined by the GBVS algorithm based on the pixel-level features corresponding to the infrared image.

[0117] The primary saliency of the target is determined based on brightness features, and / or gradient features, and / or contrast features, and / or LBP features, and / or entropy features, and / or orientation features, and / or corner density features corresponding to the infrared image.

[0118] Based on the GBVS algorithm, the pixel-level features such as brightness features, gradient features, contrast features, LBP features, entropy features, orientation features and corner density features of the target and background can be integrated to obtain the master saliency of the target.

[0119] Step S108: determining the internal saliency of the target based on the internal texture structure features corresponding to the target.

[0120] Step S110: determining the infrared camouflage effectiveness value of the target based on the main saliency and the internal saliency.

[0121] In an embodiment of the present application, determining the infrared camouflage effectiveness value of the target based on the main saliency and the internal saliency includes: calculating a comprehensive saliency value of the target based on the main saliency and the internal saliency; and determining the infrared camouflage effectiveness value of the target based on the comprehensive saliency value.

[0122] Specifically, calculating the comprehensive significance value of the target according to the main significance and the internal significance includes:

[0123] According to the main significance and the internal significance, combined with the target calculation formula, the comprehensive significance value of the target is calculated, wherein the target calculation formula includes:

[0124] S=f1*S1+f2*S2

[0125] Among them, S is the comprehensive saliency value of the target; f1 is the influence weight of S1 on the infrared camouflage effectiveness value of the target; S1 is the main saliency; f2 is the influence weight of S2 on the infrared camouflage effectiveness value of the target; S2 is the internal saliency.

[0126] In practical applications, S1 can be replaced by the mean value of the infrared saliency of the target area. Of course, because the problem studied is the evaluation of the camouflage effectiveness of infrared targets, the pixel area occupied by them in the infrared image is not as small as that of small targets. Therefore, using the average value of the infrared saliency of the camouflaged target area to represent the overall main significance level S1 of the target area is prone to evaluation result deviations due to extreme values. In order to make the evaluation results more scientific and reasonable, appropriate corrections are required. A sigmoid-like function is used as a correction function to correct the significance level of each pixel in the target area, and then the mean value of the infrared image significance of the corrected target area is calculated, so as to obtain a more objective target area image main significance level measurement S1. The form of the selected correction function is as follows:

[0127]

[0128] Among them, α is the curve curvature modulation parameter, and T is the minimum threshold value that can attract the visual attention of the human eye.

[0129] The whole modulation process is consistent with the human visual attention mechanism: before the target attracts the human eye's attention, no matter how big or small its significance level is, it cannot attract the observer to find the target; when the significance level of the target area exceeds the human eye's perception threshold, the observer may find the target and as the significance level of the target increases, the probability of the target being found increases sharply; when the significance level of the target reaches a certain level, the human eye can definitely find the target, so the increase in the significance level of the target has little effect on the probability of finding the target. Therefore, it is correct and feasible to adopt such a modulation correction method to reasonably measure the image significance level of the camouflaged target area in the infrared image. After such processing, the evaluation results of the target infrared camouflage effectiveness will be more scientific and reasonable.

[0130] Determining the infrared camouflage effectiveness value of the target based on the comprehensive saliency value includes: subtracting the comprehensive saliency value from 1 to obtain the infrared camouflage effectiveness value of the target.

[0131] In practical applications, based on the infrared images of the target and the background, the comprehensive saliency value of each pixel node in the infrared image can be obtained, and the nodes that are less similar to the surrounding neighborhood have higher image saliency. According to the camouflage principle, the essence of infrared camouflage is to make the infrared radiation characteristics of the target and the background as consistent as possible, and finally make the target and the background show a high degree of similarity in the infrared image, and achieve the most perfect fusion. Obviously, it can be considered that: if the image saliency of the infrared target is high, it means that the difference between the target and the background is large, which can easily attract visual attention, which means that its camouflage effect is poor. To this end, a conversion can be made, namely: C E =S=1-S.

[0132] Among them, C E is the infrared camouflage effectiveness value of the target, and S is the comprehensive saliency value of the target.

[0133] Based on this, the infrared camouflage effectiveness value is a specific value used to reflect the infrared camouflage effectiveness of the target. Therefore, the present application can realize the quantitative measurement and analysis of the infrared camouflage effectiveness of the target. In addition, the present application determines the infrared camouflage effectiveness value of the target based on the pixel-level features of the infrared image and the internal texture structure features of the target, so that the present application can comprehensively evaluate the infrared camouflage effectiveness of the target. The infrared camouflage effectiveness value can help relevant personnel optimize the infrared camouflage scheme of the target, which has practical significance.

[0134] By applying the infrared camouflage effectiveness evaluation method provided in the embodiment of the present application, the infrared camouflage effectiveness of the target is reflected through the final determined infrared device effectiveness value, thereby achieving quantitative measurement and analysis of the infrared camouflage effectiveness of the target.

[0135] Corresponding to the above method embodiment, this specification also provides an infrared camouflage effectiveness evaluation device embodiment, Fig.10 This is a schematic diagram of the structure of an infrared camouflage effectiveness evaluation device provided in an embodiment of the present application. Fig.10 As shown, the device comprises:

[0136] The acquisition module 1002 is configured to acquire an infrared image, wherein the infrared image carries a target and a background;

[0137] The extraction module 1004 is configured to perform feature extraction on the infrared image to generate a feature set, wherein the feature set carries pixel-level features corresponding to the infrared image and internal texture structure features corresponding to the target;

[0138] A first determination module 1006 is configured to determine the primary saliency of the target based on the pixel-level features corresponding to the infrared image;

[0139] A second determination module 1008 is configured to determine the internal saliency of the object based on the internal texture structure features corresponding to the object;

[0140] The third determination module 1010 is configured to determine the infrared camouflage effectiveness value of the target based on the main saliency and the internal saliency.

[0141] In an optional embodiment, the first determining module 1006 is further configured to:

[0142] Based on the pixel-level features corresponding to the infrared image, the primary saliency of the target is determined by using a GBVS algorithm.

[0143] In an optional embodiment, the third determining module 1010 is further configured to:

[0144] Calculating a comprehensive significance value of the target according to the main significance and the internal significance;

[0145] Based on the comprehensive saliency value, the infrared camouflage effectiveness value of the target is determined.

[0146] In an optional embodiment, the third determining module 1010 is further configured to:

[0147] According to the main significance and the internal significance, combined with the target calculation formula, the comprehensive significance value of the target is calculated, wherein the target calculation formula includes:

[0148] S=f1*S1+f2*S2

[0149] Among them, S is the comprehensive saliency value of the target; f1 is the influence weight of S1 on the infrared camouflage effectiveness value of the target; S1 is the main saliency; f2 is the influence weight of S2 on the infrared camouflage effectiveness value of the target; S2 is the internal saliency.

[0150] In an optional embodiment, the third determining module 1010 is further configured to:

[0151] The primary saliency of the target is determined based on brightness features, and / or gradient features, and / or contrast features, and / or LBP features, and / or entropy features, and / or orientation features, and / or corner density features corresponding to the infrared image.

[0152] In an optional embodiment, before the extraction module 1004, it is further configured as follows:

[0153] An infrared camouflage effectiveness evaluation index system for a target is established, wherein the infrared camouflage effectiveness evaluation index system is used to select image features for evaluating the infrared camouflage effectiveness of the target.

[0154] In an optional embodiment, the extraction module 1004 is further configured to:

[0155] The high-dimensional feature information in the feature set is converted into low-dimensional feature information.

[0156] By using the infrared camouflage effectiveness evaluation device provided in the embodiment of the present application, the infrared camouflage effectiveness of the target is reflected through the final determined infrared device effectiveness value, thereby achieving quantitative measurement and analysis of the infrared camouflage effectiveness of the target.

[0157] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the infrared camouflage effectiveness evaluation device, since it is basically similar to the infrared camouflage effectiveness evaluation method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the infrared camouflage effectiveness evaluation method embodiment.

[0158] Fig.11This is a block diagram of a computing device provided in an embodiment of the present application. The components of the computing device 1100 include but are not limited to a memory 1110 and a processor 1120. The processor 1120 is connected to the memory 1110 via a bus 1130, and the database 1150 is used to store data.

[0159] The computing device 1100 also includes an access device 1140, which enables the computing device 1100 to communicate via one or more networks 1160. Examples of these networks include a public switched telephone network (PSTN), a local area network (LAN), a wide area network (WAN), a personal area network (PAN), or a combination of communication networks such as the Internet. The access device 1140 may include one or more of any type of network interface (e.g., a network interface card (NIC)) of wired or wireless, such as an IEEE 802.11 wireless local area network (WLAN) wireless interface, a world-wide interoperability for microwave access (Wi-MAX) interface, an Ethernet interface, a universal serial bus (USB) interface, a cellular network interface, a Bluetooth interface, and a near field communication (NFC).

[0160] In one embodiment of the present specification, the above components of the computing device 1100 and Fig.11 Other components not shown in the figure may also be connected to each other, for example, via a bus. It should be understood that Fig.11 The computing device structure block diagram shown is only for the purpose of illustration, and is not intended to limit the scope of this specification. Those skilled in the art can add or replace other components as needed.

[0161] The computing device 1100 may be any type of stationary or mobile computing device, including a mobile computer or mobile computing device (e.g., a tablet computer, a personal digital assistant, a laptop computer, a notebook computer, a netbook, etc.), a mobile phone (e.g., a smart phone), a wearable computing device (e.g., a smart watch, smart glasses, etc.), or other types of mobile devices, or a stationary computing device such as a desktop computer or a personal computer (PC). The computing device 1100 may also be a mobile or stationary server.

[0162] The processor 1120 is used to execute the following computer executable instructions, which, when executed by the processor, implement the steps of the above-mentioned infrared camouflage effectiveness evaluation method.

[0163] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the computing device embodiment, since it is basically similar to the infrared camouflage effectiveness evaluation method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the infrared camouflage effectiveness evaluation method embodiment.

[0164] An embodiment of the present specification also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the steps of the above-mentioned infrared camouflage effectiveness evaluation method.

[0165] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the computer-readable storage medium embodiment, since it is basically similar to the infrared camouflage effectiveness evaluation method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the infrared camouflage effectiveness evaluation method embodiment.

[0166] An embodiment of the present specification further provides a computer program, wherein when the computer program is executed in a computer, the computer is caused to execute the steps of the above-mentioned infrared camouflage effectiveness evaluation method.

[0167] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the computer program embodiment, since it is basically similar to the infrared camouflage effectiveness evaluation method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the infrared camouflage effectiveness evaluation method embodiment.

[0168] The above is a description of a specific embodiment of the specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0169] The computer instructions include computer program codes, which may be in source code form, object code form, executable files or some intermediate forms, etc. The computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium, etc. It should be noted that the content contained in the computer-readable medium may be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electric carrier signals and telecommunication signals.

[0170] It should be noted that the above is a description of a specific embodiment of the present specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the embodiments of the present specification.

[0171] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0172] The preferred embodiments of this specification disclosed above are only used to help explain this specification. The optional embodiments do not describe all the details in detail, nor do they limit the invention to only the specific implementation methods described. Obviously, many modifications and changes can be made according to the content of the embodiments of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the embodiments of this specification, so that technicians in the relevant technical field can well understand and use this specification. This specification is only limited by the claims and their full scope and equivalents.

Claims

1. A method for evaluating the effectiveness of infrared camouflage, characterized in that: include: Acquire an infrared image, wherein the infrared image carries a target and a background; Extracting features from the infrared image to generate a feature set, wherein the feature set carries pixel-level features corresponding to the infrared image and internal texture structure features corresponding to the target; Determining the primary saliency of the target based on the pixel-level features corresponding to the infrared image; Determining the internal saliency of the target based on the internal texture structure features corresponding to the target; Based on the main saliency and the internal saliency, an infrared camouflage effectiveness value of the target is determined.

2. The method according to claim 1, characterized in that The determining the primary saliency of the target based on the pixel-level features corresponding to the infrared image includes: Based on the pixel-level features corresponding to the infrared image, the primary saliency of the target is determined by using a GBVS algorithm.

3. The method according to claim 1, characterized in that The step of determining the infrared camouflage effectiveness value of the target based on the main saliency and the internal saliency includes: Calculating a comprehensive significance value of the target according to the main significance and the internal significance; Based on the comprehensive saliency value, the infrared camouflage effectiveness value of the target is determined.

4. The method according to claim 3, characterized in that The step of calculating the comprehensive significance value of the target according to the main significance and the internal significance includes: According to the main significance and the internal significance, combined with the target calculation formula, the comprehensive significance value of the target is calculated, wherein the target calculation formula includes: S=f1*S1+f2*S2 Among them, S is the comprehensive saliency value of the target; f1 is the influence weight of S1 on the infrared camouflage effectiveness value of the target; S1 is the main saliency; f2 is the influence weight of S2 on the infrared camouflage effectiveness value of the target; S2 is the internal saliency.

5. The method according to claim 1, characterized in that The determining the primary saliency of the target based on the pixel-level features corresponding to the infrared image includes: The primary saliency of the target is determined based on brightness features, and / or gradient features, and / or contrast features, and / or LBP features, and / or entropy features, and / or orientation features, and / or corner density features corresponding to the infrared image.

6. The method according to claim 1, characterized in that The step of extracting features from the infrared image and generating a feature set includes: An infrared camouflage effectiveness evaluation index system for a target is established, wherein the infrared camouflage effectiveness evaluation index system is used to select image features for evaluating the infrared camouflage effectiveness of the target.

7. The method according to claim 1, characterized in that After extracting features from the infrared image and generating a feature set, the method includes: The high-dimensional feature information in the feature set is converted into low-dimensional feature information.

8. A method for evaluating the effectiveness of infrared camouflage, characterized in that: include: An acquisition module is configured to acquire an infrared image, wherein the infrared image carries a target and a background; An extraction module is configured to perform feature extraction on the infrared image to generate a feature set, wherein the feature set carries pixel-level features corresponding to the infrared image and internal texture structure features corresponding to the target; A first determination module is configured to determine the primary saliency of the target based on the pixel-level features corresponding to the infrared image; A second determination module is configured to determine the internal saliency of the target based on the internal texture structure features corresponding to the target; The third determination module is configured to determine the infrared camouflage effectiveness value of the target based on the main saliency and the internal saliency.

9. A computer device, characterized in that: The computer device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, and the processor implements the steps of the method according to any one of claims 1 to 7 when executing the program.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores an implementation program for information transmission, and when the program is executed by a processor, the steps of the method according to any one of claims 1 to 8 are implemented.

Images