Camouflage target identification method, storage medium and computer equipment

Abstract

The invention discloses a camouflage target identification method, a storage medium and computer equipment. The method comprises the steps that an infrared movement acquires an infrared image of a to-be-identified target object; the multispectral lens collects visible light reflected or radiated by a target object to be recognized; the metamaterial micro-nano structure spectrum modulation film layer modulates visible light collected by the multispectral lens, and the modulated visible light is detected by the low-illumination camera core to generate a multispectral image; and the image processing chip carries out image fusion on the infrared image and the multispectral image to obtain a fused image, extracts a target object contour of the to-be-identified target object based on an infrared thermal radiation signal in the fused image, and identifies whether the to-be-identified target object is a camouflage target based on the target object contour and the modulation spectral information. By fusing the multispectral low-light and infrared characteristics of the target object to be identified, the identification precision of the camouflage target is improved.

Description

Technical Field

[0001] The present application relates to the technical field of multispectral data processing, and in particular to a camouflaged target recognition method, storage medium, and computer equipment. Background Art

[0002] During combat, both enemy and friendly personnel generally wear camouflage uniforms and adopt camouflage methods when performing tasks. In the images presented by traditional low-light imaging systems, the contrast between the target and the background is low, making it difficult to detect camouflage objects with colors close to the background, which in turn leads to low accuracy in identifying camouflaged targets. Summary of the invention

[0003] In view of this, the present application provides a camouflaged target recognition method and storage medium, and computer equipment, which are applied to a multi-spectral low-light imaging system. The multi-spectral low-light imaging system includes an infrared movement, a multi-spectral movement, and an image processing chip. The multi-spectral movement includes a multi-spectral lens and a low-light camera core. The surface of the low-light camera core is coupled with a metamaterial micro-nanostructure spectral modulation film; the infrared movement obtains an infrared image of a target to be identified, wherein the infrared image includes an infrared thermal radiation signal of the target to be identified; the multi-spectral lens collects visible light reflected or radiated by the target to be identified; the metamaterial micro-nanostructure spectral modulation film The film layer modulates the visible light collected by the multispectral lens, and the modulated visible light is detected by the low-light camera core to generate a multispectral image, wherein the multispectral image includes the modulated spectrum information of the target to be identified in the preset visible light band; the image processing chip fuses the infrared image and the multispectral image to obtain a fused image, extracts the target contour of the target to be identified based on the infrared thermal radiation signal in the fused image, and identifies whether the target to be identified is a camouflaged target based on the target contour and the modulated spectrum information, wherein the fused image contains both the infrared thermal radiation signal and the modulated spectrum information. By fusing the multispectral micro-light and infrared characteristics of the target to be identified, the recognition accuracy of the camouflaged target is improved.

[0004] According to one aspect of the present application, a camouflaged target recognition method is provided, which is applied to a multi-spectral low-light imaging system, wherein the multi-spectral low-light imaging system includes an infrared core, a multi-spectral core and an image processing chip, wherein the multi-spectral core includes a multi-spectral lens and a low-light camera core, and a metamaterial micro-nanostructure spectral modulation film is coupled to the surface of the low-light camera core; the method includes: The infrared core acquires an infrared image of the target object to be identified, wherein the infrared image includes an infrared thermal radiation signal of the target object to be identified; The multispectral lens collects visible light reflected or radiated by the target to be identified; The metamaterial micro-nanostructure spectral modulation film layer modulates the visible light collected by the multi-spectral lens, and the modulated visible light is detected by the low-light camera core to generate a multi-spectral image, wherein the multi-spectral image includes the modulated spectral information of the target to be identified within a preset visible light band; The image processing chip performs image fusion on the infrared image and the multispectral image to obtain a fused image, extracts the target contour of the target object to be identified based on the infrared thermal radiation signal in the fused image, and identifies whether the target object to be identified is a camouflaged target based on the target contour and the modulated spectrum information, wherein the fused image contains both the infrared thermal radiation signal and the modulated spectrum information.

[0005] Optionally, the modulated visible light is detected by a low-light camera core to generate a multispectral image, including: When the modulated visible light is detected by the low-light camera core, the low-light camera core obtains the modulation spectrum information of the target to be identified based on the different transmittances corresponding to each group of spectral modulation channels in the metamaterial micro-nanostructure spectral modulation film layer, and generates a multi-spectral image based on the modulation spectrum information, wherein the metamaterial micro-nanostructure spectral modulation film layer includes multiple groups of spectral modulation channels, each group of spectral modulation channels is respectively set with multiple transmittances, and the visible light reflected or radiated by the target to be identified has different transmittances at the same visible light wavelength after being modulated by the spectral modulation channels of the metamaterial micro-nanostructure spectral modulation film layer, and the preset visible light band includes multiple visible light wavelengths.

[0006] Optionally, the low-light camera core obtains the modulation spectrum information of the target object to be identified based on the different transmittances corresponding to each group of spectrum modulation channels in the metamaterial micro-nanostructure spectrum modulation film layer, including: The low-light camera core forms a target linear equation group based on the calculation formula of the linear equation group, the multiple transmittances corresponding to each group of spectral modulation channels in the metamaterial micro-nanostructure spectral modulation film layer, and the grayscale value of the detector pixel of each group of spectral modulation channels. The target linear equation group is solved to obtain the modulation spectrum information of the target object to be identified, wherein the calculation formula of the linear equation group is: , O1, O2, ...O m is the detector pixel grayscale value corresponding to the spectral modulation channels of the first group to the mth group; F ij is the jth transmittance of the i-th group of spectral modulation channels, i is an integer between 1 and m, j is an integer between 1 and n; X1, X2, ..., X m is the modulation spectrum information of the target object to be identified corresponding to each of the spectral modulation channels of the first group to the mth group; N1, N2, ... N mare preset measurable noise items corresponding to the spectral modulation channels of the first group to the mth group, m is the total number of spectral modulation channels, and n is the total number of transmittance types.

[0007] Optionally, the image processing chip performs image fusion on the infrared image and the multispectral image to obtain a fused image, including: The image processing chip performs affine transformation on the coordinates of each pixel in the multispectral image based on the infrared image and the affine transformation formula to obtain the affine transformation coordinates of each pixel in the multispectral image, wherein the affine transformation coordinates include an affine transformation abscissa and an affine transformation ordinate, and the affine transformation formula is: , and are the affine transformation abscissa and ordinate of each pixel in the multispectral image, is the preset affine transformation coefficient, and are the horizontal and vertical coordinates of each pixel in the infrared image, and are the horizontal and vertical offsets of the multispectral image relative to the infrared image, respectively; The image processing chip performs image registration on the multispectral image and the infrared image according to the affine transformation coordinates of each pixel point in the multispectral image and the coordinates of each pixel point in the infrared image, and performs image fusion based on the registered multispectral image and the registered infrared image to obtain a fused image.

[0008] Optionally, the image processing chip performs image registration on the multispectral image and the infrared image according to the affine transformation coordinates of each pixel point in the multispectral image and the coordinates of each pixel point in the infrared image, including: The image processing chip calculates the registration point coordinates of each pixel in the infrared image based on the multispectral image, the pixel registration point coordinate calculation formula, the affine transformation coordinates of each pixel in the multispectral image and the coordinates of each pixel in the infrared image; The image processing chip performs image registration on the multispectral image and the infrared image based on the registration point coordinates of each pixel in the infrared image and the affine transformation coordinates of each pixel in the multispectral image, wherein the registration point coordinates include the registration point abscissa and the registration point ordinate, and the pixel registration point coordinate calculation formula is: , and are the horizontal coordinate and vertical coordinate of the registration point of each pixel in the infrared image, and are the affine transformation abscissa and ordinate of each pixel in the multispectral image, and are the horizontal and vertical scaling factors of the infrared image relative to the multispectral image, and are the horizontal and vertical offsets of the infrared image relative to the multispectral image, respectively.

[0009] Optionally, the image processing chip performs image fusion based on the multispectral image after image registration and the infrared image after image registration to obtain a fused image, including: The image processing chip extracts a first target object region and a first background region in the multispectral image after image registration, and extracts a second target object region and a second background region in the infrared image after image registration; The image processing chip performs color enhancement on the first target area to obtain a first target area color enhanced image, performs color enhancement on the second target area to obtain a second target area color enhanced image, performs contrast enhancement on the first background area to obtain a first background area contrast enhanced image, and performs contrast enhancement on the second background area to obtain a second background area contrast enhanced image; The image processing chip performs image fusion based on the first target area color enhanced image, the first background area contrast enhanced image, the second target area color enhanced image and the second background area contrast enhanced image to obtain a fused image.

[0010] Optionally, the image processing chip identifies whether the target to be identified is a camouflaged target based on the target contour and the modulation spectrum information, including: The image processing chip determines the modulation spectrum information of the target object based on the target object contour and the modulation spectrum information. When the modulation spectrum information of the target object belongs to the spectrum corresponding to natural materials, the target object to be identified is a non-camouflaged target; when the modulation spectrum information of the target object belongs to the spectrum corresponding to synthetic materials, the target object to be identified is a camouflaged target.

[0011] According to another aspect of the present application, a storage medium is provided, on which a computer program is stored, and when the program is executed by a processor, the above-mentioned camouflaged target identification method is implemented.

[0012] According to another aspect of the present application, a computer device is provided, including a storage medium, a processor, and a computer program stored on the storage medium and executable on the processor, wherein the processor implements the above-mentioned camouflaged target identification method when executing the program.

[0013] By means of the above technical scheme, the present application provides a camouflaged target recognition method, storage medium, and computer equipment, wherein the infrared core obtains the infrared image of the target to be identified; the multispectral lens collects the visible light reflected or radiated by the target to be identified; the metamaterial micro-nanostructure spectral modulation film layer modulates the visible light collected by the multispectral lens, and the modulated visible light is detected by the low-light camera core to generate a multispectral image; the image processing chip fuses the infrared image and the multispectral image to obtain a fused image, extracts the target contour of the target to be identified based on the infrared thermal radiation signal in the fused image, and identifies whether the target to be identified is a camouflaged target based on the target contour and the modulated spectrum information. By fusing the multispectral micro-light and infrared characteristics of the target to be identified, the recognition accuracy of the camouflaged target is improved.

[0014] 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, it can be implemented in accordance with the contents of the specification. 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

[0015] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings: Figure 1 A schematic diagram of a process of identifying a camouflaged target provided by an embodiment of the present application is shown; Figure 2 A schematic diagram of a multi-spectral lens provided in an embodiment of the present application is shown; Figure 3 A schematic diagram of a process flow of another camouflaged target identification method provided in an embodiment of the present application is shown; Figure 4 A schematic diagram of a metamaterial micro-nanostructure spectral modulation film layer provided in an embodiment of the present application is shown; Figure 5 A schematic diagram of a unit structure of a metamaterial micro-nanostructure spectral modulation film layer provided in an embodiment of the present application is shown; Figure 6 A multispectral image provided by an embodiment of the present application is shown; Figure 7 Another multispectral image provided by an embodiment of the present application is shown; Figure 8 A schematic flow chart of another camouflaged target identification method provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0016] The present application will be described in detail below with reference to the accompanying drawings and in combination with embodiments. It should be noted that the embodiments and features in the embodiments of the present application can be combined with each other without conflict.

[0017] In this embodiment, a camouflaged target recognition method is provided. Figure 1 As shown, it is applied to a multi-spectral low-light imaging system, the multi-spectral low-light imaging system includes an infrared core, a multi-spectral core and an image processing chip, the multi-spectral core includes a multi-spectral lens and a low-light camera core, and the surface of the low-light camera core is coupled with a metamaterial micro-nanostructure spectral modulation film; the method includes: Step 101: The infrared core acquires an infrared image of a target object to be identified, wherein the infrared image includes an infrared thermal radiation signal of the target object to be identified.

[0018] In the above-mentioned embodiment of the present application, it is applied to a multi-spectral low-light imaging system, and is applied to a multi-spectral low-light imaging system. The multi-spectral low-light imaging system includes an infrared movement, a multi-spectral movement and an image processing chip. The multi-spectral movement includes a multi-spectral lens and a low-light camera core. The surface of the low-light camera core is coupled with a metamaterial micro-nanostructure spectral modulation film. In particular, with respect to the infrared movement, the infrared movement includes an infrared lens and an infrared imaging device, which are used to jointly obtain an infrared image of a target object to be identified. The infrared image includes an infrared thermal radiation signal of the target object to be identified, in preparation for subsequent camouflage target identification. With respect to the low-light camera core, it can be realized by designing a domestic P3220 low-light imaging device for overall packaging, and the multi-spectral lens is designed with a metamaterial micro-nanostructure spectral modulation film layer according to the parameters of the domestic P3220 low-light imaging device. Specifically, the parameters of the domestic P3220 low-light imaging device are as shown in Table 1: Table 1

[0019] When designing a multispectral lens, i.e. a multispectral objective lens, the focal length of the multispectral objective lens (multispectral lens) is first determined based on the target size and field of view of the selected imaging device (P3220 domestic low-light imaging device). Specifically, the design is performed according to the following formula: , , Where, 2 Indicates the horizontal field of view of the product (40°), 2 Indicates the vertical field of view of the product (30°), Indicates horizontal distortion (-0.7%), Indicates vertical distortion (-0.3%), L indicates the horizontal size of the MCCD (low illumination camera core) target surface (13.3mm), H indicates the vertical size of the MCCD (low illumination camera core) target surface (9.98mm), Indicates the focal length of the multispectral lens. Without considering distortion, in order to simultaneously meet the requirements of horizontal field of view not less than 40° and vertical field of view not less than 30°, the focal length of the main field of view multispectral lens should be 18.4mm at most. Therefore, the designed multispectral lens focal length can be 18mm, the designed field of view can be 40.8°×31°, the diagonal distortion can be -1.4%, and the calculated diagonal field of view can be 50.2°. The final multispectral lens can be realized with 6 lenses, for example Figure 2 shown.

[0020] Step 102: The multispectral lens collects visible light reflected or radiated by the target object to be identified.

[0021] Step 103: The metamaterial micro-nanostructure spectral modulation film layer modulates the visible light collected by the multi-spectral lens, and the modulated visible light is detected by a low-light camera core to generate a multi-spectral image, wherein the multi-spectral image includes the modulated spectral information of the target object to be identified within a preset visible light band.

[0022] Next, the multispectral lens collects the visible light reflected or radiated by the target to be identified, and the low-light camera core modulates the visible light collected by the multispectral lens. The modulated visible light is detected by the low-light camera core to generate a multispectral image, which continues to prepare for the subsequent identification of camouflaged targets. The multispectral image includes the modulated spectrum information of the target to be identified in a preset visible light band. The preset visible light wavelength band is, for example, a wavelength band of 400 to 900nm (nanometers). Since the visible wavelength band of 400 to 900nm can contain multiple visible light wavelengths, by widening the visible light wavelength band to be identified, it is possible to better "identify" the visible light modulation spectrum information of the target to be identified at various wavelengths, which is conducive to improving the recognition accuracy.

[0023] Step 104: the image processing chip fuses the infrared image and the multi-spectral image to obtain a fused image, extracts the target contour of the target object to be identified based on the infrared thermal radiation signal in the fused image, and identifies whether the target object to be identified is a camouflaged target based on the target contour and the modulated spectrum information, wherein the fused image contains both the infrared thermal radiation signal and the modulated spectrum information.

[0024] Finally, the image processing chip fuses the infrared image and the multispectral image to obtain a fused image, extracts the target contour of the target to be identified based on the infrared thermal radiation signal in the fused image, and identifies whether the target to be identified is a camouflaged target based on the target contour and the modulated spectrum information. To this end, the image processing chip can enhance the "detection effect" of the target to be identified and improve the target detection capability by fusing the infrared image with the multispectral image, and then performs feature variable importance analysis and feature selection on the fused image, and classifies and evaluates the accuracy of the selected features, so as to extract the target contour according to the classification effect, thereby realizing camouflaged target recognition and improving the recognition accuracy of camouflaged targets.

[0025] By applying the technical solution of this embodiment, the infrared core acquires the infrared image of the target to be identified; the multispectral lens collects the visible light reflected or radiated by the target to be identified; the metamaterial micro-nanostructure spectral modulation film layer modulates the visible light collected by the multispectral lens, and the modulated visible light is detected by the low-light camera core to generate a multispectral image; the image processing chip fuses the infrared image and the multispectral image to obtain a fused image, extracts the target contour of the target to be identified based on the infrared thermal radiation signal in the fused image, and identifies whether the target to be identified is a camouflaged target based on the target contour and the modulated spectrum information. By fusing the multispectral micro-light and infrared characteristics of the target to be identified, the recognition accuracy of the camouflaged target is improved.

[0026] Further, as a refinement and expansion of the specific implementation of the above embodiment, in order to fully illustrate the specific implementation process of this embodiment, another camouflaged target recognition method is provided, which is applied to a multi-spectral low-light imaging system, wherein the multi-spectral low-light imaging system includes an infrared core, a multi-spectral core and an image processing chip, wherein the multi-spectral core includes a multi-spectral lens and a low-light camera core, and a metamaterial micro-nanostructure spectral modulation film is coupled to the surface of the low-light camera core; Figure 3 As shown, the method includes: Step 201: The infrared core acquires an infrared image of a target object to be identified, wherein the infrared image includes an infrared thermal radiation signal of the target object to be identified.

[0027] Step 202: The multispectral lens collects visible light reflected or radiated by the target object to be identified.

[0028] Step 203: The metamaterial micro-nanostructure spectrum modulation film layer modulates the visible light collected by the multi-spectral lens.

[0029] In the above embodiment of the present application, the infrared movement obtains an infrared image of the target to be identified, wherein the infrared image includes the infrared thermal radiation signal of the target to be identified. Next, the multi-spectral lens collects the visible light reflected or radiated by the target to be identified, and the metamaterial micro-nanostructure spectral modulation film layer modulates the visible light collected by the multi-spectral lens. Since the metamaterial micro-nanostructure spectral modulation film layer includes multiple groups of spectral modulation channels, each group of spectral modulation channels is respectively provided with multiple transmittances. When the visible light passes through the metamaterial micro-nanostructure spectral modulation film layer, the visible light will have different transmittances at the same visible light wavelength. The metamaterial micro-nanostructure spectral modulation film layer, for example Figure 4 As shown, Figure 4 In the above figure, each group of spectral modulation channels is set with 9 transmittances, and 8 transmittances can be set according to actual requirements. The metamaterial micro-nanostructure spectral modulation film layer adopts a symmetrical C-shaped metal metamaterial three-dimensional structure, and its unit structure is as follows: Figure 5 As shown, since the molecular or atomic structure of each substance is specific, each substance has its own unique spectral absorption or emission mode, so artificial materials and natural materials will have different absorption or emission spectra at different wavelengths. Each substance has a characteristic spectral pattern, similar to a human fingerprint, so it is called a "fingerprint" characteristic. Visible light modulated by the metamaterial micro-nanostructure spectral modulation film still has this fingerprint characteristic, so different substances can be identified and distinguished.

[0030] Step 204, when the modulated visible light is detected by the low-light camera core, the low-light camera core constructs a target linear equation group based on the linear equation group calculation formula, the multiple transmittances corresponding to each group of spectral modulation channels in the metamaterial micro-nanostructure spectral modulation film layer, and the detector pixel grayscale value of each group of spectral modulation channels, and solves the target linear equation group to obtain the modulation spectrum information of the target object to be identified, wherein the linear equation group calculation formula is: , O1, O2, ...O m is the detector pixel grayscale value corresponding to the spectral modulation channels of the first group to the mth group; F ij is the jth transmittance of the i-th group of spectral modulation channels, i is an integer between 1 and m, j is an integer between 1 and n; X1, X2, ..., X m is the modulation spectrum information of the target object to be identified corresponding to each of the spectral modulation channels of the first group to the mth group; N1, N2, ... N m are preset measurable noise items corresponding to the spectral modulation channels of the first group to the mth group, m is the total number of spectral modulation channels, and n is the total number of transmittance types.

[0031] Next, the low-light camera core obtains the modulation spectrum information of the target to be identified based on the calculation formula of the linear equation group. Specifically, when each group of spectral modulation channels is set to nine channels, due to the transmittance of each group of nine channels, the grayscale values ​​of the detector pixels corresponding to each group of spectral modulation channels can form a nine-variable linear equation group. The solution of this equation group is the intensity value on the nine spectral bands. By traversing the entire detector image, the multispectral image values ​​of the nine spectral bands of the target to be identified can be obtained, that is, the modulation spectrum information of the target to be identified. The target to be identified is in the wavelength range of 400 to 900nm, and the situations of various wavelengths and various transmittances are as follows. Figure 6 As shown in the figure, the various wavelengths and various transmittances after modulation are as follows Figure 7 shown.

[0032] Step 205: The low-light camera core generates a multi-spectral image based on the modulated spectral information, wherein the metamaterial micro-nanostructure spectral modulation film layer includes multiple groups of spectral modulation channels, each group of spectral modulation channels is respectively set with multiple transmittances, and the visible light reflected or radiated by the target object to be identified has different transmittances at the same visible light wavelength after being modulated by the spectral modulation channels of the metamaterial micro-nanostructure spectral modulation film layer, the preset visible light band includes multiple visible light wavelengths, and the multi-spectral image includes the modulated spectral information of the target object to be identified within the preset visible light band.

[0033] Step 206: The image processing chip performs image fusion on the infrared image and the multispectral image to obtain a fused image, and extracts the object contour of the target object to be identified based on the infrared thermal radiation signal in the fused image.

[0034] Step 207, the image processing chip determines the modulation spectrum information of the target object based on the target object contour and the modulation spectrum information. When the modulation spectrum information of the target object belongs to the spectrum corresponding to natural materials, the target object to be identified is a non-camouflaged target. When the modulation spectrum information of the target object belongs to the spectrum corresponding to synthetic materials, the target object to be identified is a camouflaged target, wherein the fused image contains both infrared thermal radiation signals and modulation spectrum information.

[0035] Next, the low-light camera core generates a multispectral image based on the modulation spectrum information, and the image processing chip fuses the infrared image and the multispectral image to obtain a fused image, and extracts the target contour of the target to be identified based on the infrared thermal radiation signal in the fused image. The image processing chip determines the modulation spectrum information of the target based on the target contour and the modulation spectrum information. When the modulation spectrum information of the target belongs to the spectrum corresponding to the natural material, the target to be identified is a non-camouflaged target. When the modulation spectrum information of the target belongs to the spectrum corresponding to the synthetic material, the target to be identified is a camouflaged target, wherein the fused image contains both the infrared thermal radiation signal and the modulation spectrum information. It is possible to use the characteristics of different materials with different characteristic spectral patterns to distinguish whether the target belongs to a natural material or a synthetic material. At the same time, the multispectral image modulated by the metamaterial micro-nanostructure spectral modulation film "enhances" the spectral characteristics of the target to be identified at various wavelengths, which is conducive to improving the recognition accuracy.

[0036] By applying the technical solution of this embodiment, the infrared core obtains an infrared image of the target to be identified, the multispectral lens collects visible light reflected or radiated by the target to be identified, and the metamaterial micro-nanostructure spectral modulation film layer modulates the visible light. When the modulated visible light is detected by the low-light camera core, the low-light camera core forms a target linear equation group based on the linear equation group calculation formula, solves the target linear equation group, and obtains the modulation spectrum information of the target to be identified. The low-light camera core generates a multispectral image based on the modulation spectrum information. The image processing chip fuses the infrared image and the multispectral image to obtain a fused image. The target contour of the target to be identified is extracted based on the infrared thermal radiation signal in the fused image. The image processing chip determines the modulation spectrum information of the target based on the target contour and the modulation spectrum information. When the modulation spectrum information of the target belongs to the spectrum corresponding to the natural material, the target to be identified is a non-camouflaged target. When the modulation spectrum information of the target belongs to the spectrum corresponding to the synthetic material, the target to be identified is a camouflaged target. By fusing the multispectral micro-light and infrared characteristics of the target to be identified, the recognition accuracy of the camouflaged target is improved.

[0037] Further, as a refinement and expansion of the specific implementation of the above embodiment, in order to fully illustrate the specific implementation process of this embodiment, another camouflaged target recognition method is provided, which is applied to a multi-spectral low-light imaging system, wherein the multi-spectral low-light imaging system includes an infrared core, a multi-spectral core and an image processing chip, wherein the multi-spectral core includes a multi-spectral lens and a low-light camera core, and a metamaterial micro-nanostructure spectral modulation film is coupled to the surface of the low-light camera core; Figure 8 As shown, the method includes: Step 301: The infrared core acquires an infrared image of a target object to be identified, wherein the infrared image includes an infrared thermal radiation signal of the target object to be identified.

[0038] Step 302: The multispectral lens collects visible light reflected or radiated by the target object to be identified.

[0039] Step 303: The metamaterial micro-nanostructure spectral modulation film layer modulates the visible light collected by the multi-spectral lens, and the modulated visible light is detected by a low-light camera core to generate a multi-spectral image, wherein the multi-spectral image includes the modulated spectral information of the target object to be identified within a preset visible light band.

[0040] In the above-mentioned embodiment of the present application, the infrared movement obtains an infrared image of the target to be identified, which is convenient for the subsequent "positioning" of the target to be identified. The multispectral lens collects the visible light reflected or radiated by the target to be identified. The material micro-nanostructure spectral modulation film layer modulates the visible light collected by the multispectral lens. The modulated visible light is detected by the low-light camera core to generate a multispectral image, wherein the multispectral image includes the modulated spectral information of the target to be identified within a preset visible light band. The preset visible light wavelength band is, for example, 400~900nm. The modulated spectral information within the "wide" wavelength band can improve the recognition accuracy for subsequent camouflaged targets.

[0041] In step 304, the image processing chip performs affine transformation on the coordinates of each pixel point in the multispectral image based on the infrared image and the affine transformation formula to obtain the affine transformation coordinates of each pixel point in the multispectral image, wherein the affine transformation coordinates include an affine transformation abscissa and an affine transformation ordinate, and the affine transformation formula is: , and are the affine transformation abscissa and ordinate of each pixel in the multispectral image, is the preset affine transformation coefficient, and are the horizontal and vertical coordinates of each pixel in the infrared image, and are the horizontal and vertical offsets of the multispectral image relative to the infrared image, respectively.

[0042] Next, it is necessary to fuse the infrared image and the multispectral image. Before image fusion, image registration is required first. Image registration is the core technology of image fusion, which directly determines the quality of the fusion result. Direct fusion of unregistered images will produce false information and cause misleading recognition. Since the physical properties reflected by infrared images and multispectral images are different, the grayscale correlation between the two is small. If the imaging optical axes are not parallel or the resolutions are different, the imaging targets of different sensors will produce differences or distortions at the same position, such as translation and rotation. The first step of image registration is to perform affine transformation. For the image information of the same scene, since the number of pixels in the infrared image is less than that in the multispectral image, each pixel of the infrared image contains the information of up to four adjacent multispectral image pixels. Therefore, the affine transformation of the infrared image and the multispectral image is very necessary. Specifically, the multispectral core takes the infrared image as the reference, and performs affine transformation on the coordinates of each pixel in the multispectral image based on the affine transformation formula to obtain the affine transformation coordinates of each pixel in the multispectral image. For this reason, the affine transformation coordinates of each pixel of the multispectral image relative to the infrared image are obtained.

[0043] Step 305, the image processing chip calculates the registration point coordinates of each pixel in the infrared image based on the multispectral image, according to the pixel registration point coordinate calculation formula, the affine transformation coordinates of each pixel in the multispectral image and the coordinates of each pixel in the infrared image.

[0044] Step 306, the image processing chip performs image registration on the multispectral image and the infrared image based on the registration point coordinates of each pixel in the infrared image and the affine transformation coordinates of each pixel in the multispectral image, wherein the registration point coordinates include the registration point abscissa and the registration point ordinate, and the pixel registration point coordinate calculation formula is: , and are the horizontal coordinate and vertical coordinate of the registration point of each pixel in the infrared image, and are the affine transformation abscissa and ordinate of each pixel in the multispectral image, and are the horizontal and vertical scaling factors of the infrared image relative to the multispectral image, and are the horizontal and vertical offsets of the infrared image relative to the multispectral image, respectively.

[0045] Next, the multispectral core calculates the registration point coordinates of each pixel in the infrared image based on the multispectral image as a reference, according to the pixel registration point coordinate calculation formula, the affine transformation coordinates of each pixel in the multispectral image and the coordinates of each pixel in the infrared image. The multispectral core performs image registration on the multispectral image and the infrared image based on the registration point coordinates of each pixel in the infrared image and the affine transformation coordinates of each pixel in the multispectral image.

[0046] Specifically, a target that is visible in both infrared and multispectral images can be designed and manufactured, and its shape is rectangular. Start the calibration procedure, that is, the calibration order is infrared system (infrared image) first, then low light system (multispectral image). Since the resolution of infrared image is low, the multispectral image is used as the reference image to obtain the offset of the infrared image relative to the multispectral image coordinates. and , the scaling factor in the horizontal and vertical directions and Next, save and , and , for a fusion scene with a fixed optical system, and , and Recalculate the coordinates of the infrared image according to the stored parameters, assuming that the pixel coordinates of the multispectral image are , the pixel coordinates of the infrared image are , then the registered coordinates of the corresponding points in the infrared image can be expressed as Therefore, the purpose of adaptive registration in the image fusion process is achieved. When the optical system changes, the re-registration can be completed by repeating the operation.

[0047] In step 307 , the image processing chip extracts the first target object region and the first background region in the multispectral image after image registration, and extracts the second target object region and the second background region in the infrared image after image registration.

[0048] In step 308, the image processing chip performs color enhancement on the first target area to obtain a first target area color enhanced image, performs color enhancement on the second target area to obtain a second target area color enhanced image, performs contrast enhancement on the first background area to obtain a first background area contrast enhanced image, and performs contrast enhancement on the second background area to obtain a second background area contrast enhanced image.

[0049] Step 309: The image processing chip performs image fusion based on the first target area color enhanced image, the first background area contrast enhanced image, the second target area color enhanced image, and the second background area contrast enhanced image to obtain a fused image.

[0050] Next, the multispectral movement selects the multispectral image after image registration or the infrared image after image registration as the image to be fused, extracts the target area and the background area in the image to be fused, performs color enhancement on the target area to obtain a target area color enhanced image, and performs contrast enhancement on the background area to obtain a background area contrast enhanced image, and performs image fusion based on the first target area color enhanced image and the first background area contrast enhanced image obtained from the multispectral image after image registration, and the second target area color enhanced image and the second background area contrast enhanced image obtained from the infrared image after image registration to obtain a fused image.

[0051] Specifically, after completing the image registration, the image fusion operation can be started, in which target detection is the main application purpose of image fusion technology. In order to enhance the target, the target segmentation method can be used, that is, the target to be identified is extracted from the source image, and the grayscale target image (the first target area and the second target area) is mapped to the RGB channel to obtain a target enhanced color image (the first target area color enhanced image and the second target area color enhanced image). In order to optimize the background image contrast, the brightness contrast optimization method can be used to place the Laplace fusion image of the dual-channel background into the brightness component of the lαβ space to obtain a contrast-optimized background fusion image. Finally, the target image and the background image are combined to obtain a color fusion image with natural color, contrast and target enhancement that conforms to human visual observation. For example, the image is segmented based on saliency to obtain a dual-channel target and background image; the target part, that is, the target area, is colorized and enhanced, and the background part is optimized for brightness contrast. The Laplace fused image of the dual-channel background is placed in channel 1, and the α channel and β channel are color transferred, and then converted from the lαβ space to the RGB space; the target color fused image and the background color fused image are combined to finally form a color fused image (fused image) with target enhancement and contrast optimization.

[0052] In step 310, the image processing chip extracts the contour of the target object to be identified based on the infrared thermal radiation signal in the fused image, and identifies whether the target object to be identified is a camouflaged target based on the target contour and the modulation spectrum information, wherein the fused image contains both the infrared thermal radiation signal and the modulation spectrum information.

[0053] Finally, the image processing chip extracts the target contour of the target to be identified based on the infrared thermal radiation signal in the fused image, and identifies whether the target to be identified is a camouflaged target based on the target contour and the modulation spectrum information. High-quality imaging can be achieved, making up for the defect that existing night vision technology is difficult to identify camouflaged targets. By applying the technical solution of this embodiment, the infrared core acquires the infrared image of the target to be identified, the multispectral lens collects the visible light reflected or radiated by the target to be identified, the metamaterial micro-nanostructure spectral modulation film layer modulates the visible light collected by the multispectral lens, and the modulated visible light is detected by the low-light camera core to generate a multispectral image. The image processing chip performs affine transformation on the coordinates of each pixel point in the multispectral image based on the infrared image and the affine transformation formula to obtain the affine transformation coordinates of each pixel point in the multispectral image. The image processing chip calculates the registration point coordinates of each pixel point in the infrared image based on the pixel registration point coordinate calculation formula, the affine transformation coordinates of each pixel point in the multispectral image, and the coordinates of each pixel point in the infrared image. The image processing chip performs image registration on the multispectral image and the infrared image based on the registration point coordinates of each pixel point in the infrared image and the affine transformation coordinates of each pixel point in the multispectral image. The image processing chip extracts the first target object area and the first background area in the multispectral image after image registration, and extracts the second target object area and the second background area in the infrared image after image registration. The image processing chip performs color enhancement on the first target area to obtain a color enhanced image of the first target area, performs color enhancement on the second target area to obtain a color enhanced image of the second target area, performs contrast enhancement on the first background area to obtain a contrast enhanced image of the first background area, and performs contrast enhancement on the second background area to obtain a contrast enhanced image of the second background area. The image processing chip performs image fusion based on the color enhanced image of the first target area, the contrast enhanced image of the first background area, the color enhanced image of the second target area, and the contrast enhanced image of the second background area to obtain a fused image. The recognition accuracy of the camouflaged target is improved.

[0054] Based on the above Figure 1 , Figure 3 and Figure 8 The method shown in the embodiment of the present application is accordingly provided with a storage medium on which a computer program is stored, and when the computer program is executed by a processor, the above-mentioned Figure 1 , Figure 3 and Figure 8 The camouflaged target recognition method shown.

[0055] Based on this understanding, the technical solution of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), and includes a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each implementation scenario of the present application.

[0056] Based on the above Figure 1 , Figure 3 and Figure 8 In order to achieve the above-mentioned purpose, the embodiment of the present application also provides a computer device, which can be a personal computer, a server, a network device, etc. The computer device includes a storage medium and a processor; the storage medium is used to store a computer program; the processor is used to execute the computer program to achieve the above-mentioned Figure 1 , Figure 3 and Figure 8 The camouflaged target recognition method shown.

[0057] Optionally, the computer device may also include a user interface, a network interface, a camera, a radio frequency (RF) circuit, a sensor, an audio circuit, a WI-FI module, etc. The user interface may include a display, an input unit such as a keyboard, etc., and the optional user interface may also include a USB interface, a card reader interface, etc. The network interface may optionally include a standard wired interface, a wireless interface (such as a Bluetooth interface, a WI-FI interface), etc.

[0058] Those skilled in the art will appreciate that the computer device structure provided in this embodiment does not limit the computer device, and may include more or fewer components, or integrate certain components, or arrange the components differently.

[0059] The storage medium may also include an operating system and a network communication module. The operating system is a program that manages and saves the hardware and software resources of the computer device, and supports the operation of information processing programs and other software and / or programs. The network communication module is used to realize the communication between the various software inside the storage medium, and the communication with other hardware and software in the physical device.

[0060] Through the description of the above implementation methods, those skilled in the art can clearly understand that the present application can be implemented by means of software plus necessary general hardware platforms, or by hardware to implement an infrared movement to obtain an infrared image of a target to be identified; a multi-spectral lens collects visible light reflected or radiated by the target to be identified; a metamaterial micro-nanostructure spectral modulation film layer modulates the visible light collected by the multi-spectral lens, and the modulated visible light is detected by a low-light camera core to generate a multi-spectral image; an image processing chip fuses the infrared image and the multi-spectral image to obtain a fused image, extracts the target contour of the target to be identified based on the infrared thermal radiation signal in the fused image, and identifies whether the target to be identified is a camouflaged target based on the target contour and the modulated spectrum information. By fusing the multi-spectral micro-light and infrared characteristics of the target to be identified, the recognition accuracy of the camouflaged target is improved.

[0061] The above serial numbers of this application are only for description and do not represent the advantages and disadvantages of the implementation scenarios. The above disclosure is only a few specific implementation scenarios of this application, but this application is not limited to them, and any changes that can be thought of by technicians in this field should fall within the scope of protection of this application.

Claims

1. A camouflaged target recognition method, characterized in that: Applied to a multi-spectral low-light imaging system, the multi-spectral low-light imaging system comprises an infrared core, a multi-spectral core and an image processing chip, the multi-spectral core comprises a multi-spectral lens and a low-light camera core, and the surface of the low-light camera core is coupled with a metamaterial micro-nanostructure spectral modulation film; The method comprises: The infrared core acquires an infrared image of the target object to be identified, wherein the infrared image includes an infrared thermal radiation signal of the target object to be identified; The multispectral lens collects visible light reflected or radiated by the target to be identified; The metamaterial micro-nanostructure spectral modulation film layer modulates the visible light collected by the multi-spectral lens, and the modulated visible light is detected by the low-light camera core to generate a multi-spectral image, wherein the multi-spectral image includes the modulated spectral information of the target to be identified within a preset visible light band; The image processing chip performs image fusion on the infrared image and the multispectral image to obtain a fused image, extracts the target contour of the target object to be identified based on the infrared thermal radiation signal in the fused image, and identifies whether the target object to be identified is a camouflaged target based on the target contour and the modulated spectrum information, wherein the fused image contains both the infrared thermal radiation signal and the modulated spectrum information.

2. The method according to claim 1, characterized in that The modulated visible light is detected by a low-light camera core to generate a multispectral image, including: When the modulated visible light is detected by the low-light camera core, the low-light camera core obtains the modulation spectrum information of the target to be identified based on the different transmittances corresponding to each group of spectral modulation channels in the metamaterial micro-nanostructure spectral modulation film layer, and generates a multi-spectral image based on the modulation spectrum information, wherein the metamaterial micro-nanostructure spectral modulation film layer includes multiple groups of spectral modulation channels, each group of spectral modulation channels is respectively set with multiple transmittances, and the visible light reflected or radiated by the target to be identified has different transmittances at the same visible light wavelength after being modulated by the spectral modulation channels of the metamaterial micro-nanostructure spectral modulation film layer, and the preset visible light band includes multiple visible light wavelengths.

3. The method according to claim 2, characterized in that The low-light camera core obtains the modulation spectrum information of the target object to be identified based on the different transmittances corresponding to each group of spectrum modulation channels in the metamaterial micro-nanostructure spectrum modulation film layer, including: The low-light camera core forms a target linear equation group based on the calculation formula of the linear equation group, the multiple transmittances corresponding to each group of spectral modulation channels in the metamaterial micro-nanostructure spectral modulation film layer, and the grayscale value of the detector pixel of each group of spectral modulation channels. The target linear equation group is solved to obtain the modulation spectrum information of the target object to be identified, wherein the calculation formula of the linear equation group is: ， O1, O2, ...O m is the detector pixel grayscale value corresponding to the spectral modulation channels of the first group to the mth group; F ij is the jth transmittance of the i-th group of spectral modulation channels, i is an integer between 1 and m, j is an integer between 1 and n; X1, X2, ..., X m is the modulation spectrum information of the target object to be identified corresponding to each of the spectral modulation channels of the first group to the mth group; N1, N2, ... N m are preset measurable noise items corresponding to the spectral modulation channels of the first group to the mth group, m is the total number of spectral modulation channels, and n is the total number of transmittance types.

4. The method according to claim 1, characterized in that: The image processing chip performs image fusion on the infrared image and the multispectral image to obtain a fused image, including: The image processing chip performs affine transformation on the coordinates of each pixel in the multispectral image based on the infrared image and the affine transformation formula to obtain the affine transformation coordinates of each pixel in the multispectral image, wherein the affine transformation coordinates include an affine transformation abscissa and an affine transformation ordinate, and the affine transformation formula is: ， and are the affine transformation abscissa and ordinate of each pixel in the multispectral image, is the preset affine transformation coefficient, and are the horizontal and vertical coordinates of each pixel in the infrared image, and are the horizontal and vertical offsets of the multispectral image relative to the infrared image, respectively; The image processing chip performs image registration on the multispectral image and the infrared image according to the affine transformation coordinates of each pixel point in the multispectral image and the coordinates of each pixel point in the infrared image, and performs image fusion based on the registered multispectral image and the registered infrared image to obtain a fused image.

5. The method according to claim 4, characterized in that The image processing chip performs image registration on the multispectral image and the infrared image according to the affine transformation coordinates of each pixel point in the multispectral image and the coordinates of each pixel point in the infrared image, including: The image processing chip calculates the registration point coordinates of each pixel in the infrared image based on the multispectral image, the pixel registration point coordinate calculation formula, the affine transformation coordinates of each pixel in the multispectral image and the coordinates of each pixel in the infrared image; The image processing chip performs image registration on the multispectral image and the infrared image based on the registration point coordinates of each pixel in the infrared image and the affine transformation coordinates of each pixel in the multispectral image, wherein the registration point coordinates include the registration point abscissa and the registration point ordinate, and the pixel registration point coordinate calculation formula is: ， and are the horizontal and vertical coordinates of the registration points of each pixel in the infrared image, and are the affine transformation abscissa and ordinate of each pixel in the multispectral image, and are the horizontal and vertical scaling factors of the infrared image relative to the multispectral image, and are the horizontal and vertical offsets of the infrared image relative to the multispectral image, respectively.

6. The method according to claim 4, characterized in that The image processing chip performs image fusion based on the multispectral image after image registration and the infrared image after image registration to obtain a fused image, including: The image processing chip extracts a first target object region and a first background region in the multispectral image after image registration, and extracts a second target object region and a second background region in the infrared image after image registration; The image processing chip performs color enhancement on the first target area to obtain a first target area color enhanced image, performs color enhancement on the second target area to obtain a second target area color enhanced image, performs contrast enhancement on the first background area to obtain a first background area contrast enhanced image, and performs contrast enhancement on the second background area to obtain a second background area contrast enhanced image; The image processing chip performs image fusion based on the first target area color enhanced image, the first background area contrast enhanced image, the second target area color enhanced image and the second background area contrast enhanced image to obtain a fused image.

7. The method according to any one of claims 1 to 6, characterized in that The image processing chip identifies whether the target to be identified is a camouflaged target based on the target contour and the modulation spectrum information, including: The image processing chip determines the modulation spectrum information of the target object based on the target object contour and the modulation spectrum information. When the modulation spectrum information of the target object belongs to the spectrum corresponding to natural materials, the target object to be identified is a non-camouflaged target; when the modulation spectrum information of the target object belongs to the spectrum corresponding to synthetic materials, the target object to be identified is a camouflaged target.

8. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for identifying a camouflaged target as claimed in any one of claims 1 to 7 is implemented.

9. A computer device comprising a storage medium, a processor, and a computer program stored in the storage medium and executable on the processor, characterized in that: When the processor executes the computer program, the method for identifying a camouflaged target according to any one of claims 1 to 7 is implemented.

Images