Regional image texture complexity calculation method and device based on IVE and storage medium

Abstract

The invention provides an IVE-based regional image texture complexity calculation method and device and a storage medium, and the method comprises the steps: obtaining a to-be-processed RGB image, determining a region of interest according to the RGB image, and obtaining the coordinate information of the region of interest; carrying and storing the region of interest by using a dma operator of an IVE hardware acceleration module according to the coordinate information of the region of interest to obtain a region image; calculating a structure tensor of the regional image through an image processing operator in the IVE hardware acceleration module to obtain a class structure tensor image; performing binarization processing on the class structure tensor image to obtain a binarized image of the class structure tensor image; and carrying out quantity statistics on the maximum value according to the binary image of the class structure tensor image, and calculating to obtain the texture complexity. According to the calculation method, on the premise that the performance and the effect are guaranteed, the calculation amount is greatly reduced, and the calculation of the texture complexity of the regional image is accelerated.

Description

Technical Field

[0001] The present application relates to the field of image processing technology, and in particular to a method, device and storage medium for calculating regional image texture complexity based on IVE. Background Art

[0002] In the traditional frame-selected target tracking process, the user usually manually selects the ROI (Region of Interest) for the target image in the first frame. There is a problem that the ROI does not contain the target or the texture is not complex enough, resulting in reduced tracking performance. Therefore, it is necessary to judge the target or texture complexity of the ROI in the first frame to ensure tracking performance.

[0003] For the texture complexity calculation of the ROI of the first frame, the existing technology usually calculates the texture complexity by calculating the ROI structure tensor based on the image processing library such as OPEN CV on the CPU, and a large number of mathematical operations are required in this process. However, the image operation based on the CPU is not only slow and inefficient, but also the performance will be further reduced as the image size increases. Summary of the invention

[0004] In order to solve the technical problem of low efficiency of image texture complexity calculation based on CPU in the prior art, the present application proposes a method, device and storage medium for calculating regional image texture complexity based on IVE.

[0005] According to the first aspect of the present application, a method for calculating regional image texture complexity based on IVE is proposed, comprising: Acquire an RGB image to be processed, determine a region of interest according to the RGB image, and acquire coordinate information of the region of interest; According to the coordinate information of the region of interest, the region of interest is moved and saved using the DMA operator of the IVE hardware acceleration module to obtain a regional image; Calculating the structure tensor of the regional image by using the image processing operator in the IVE hardware acceleration module to obtain a quasi-structure tensor image; Binarizing the class structure tensor image to obtain a binary image of the class structure tensor image; According to the binarized image of the class structure tensor image, the maximum value is quantitatively counted to calculate the texture complexity.

[0006] Preferably, calculating the structure tensor of the regional image by an image processing operator in the IVE hardware acceleration module to obtain a quasi-structure tensor image includes: Performing grayscale processing on the regional image to obtain a grayscale image; Performing a convolution operation on the grayscale image based on a Sobel operator to obtain first gradient images in horizontal and vertical directions respectively; quantizing and converting the two first gradient images into absolute values ​​to obtain a second gradient image; A weighted superposition calculation is performed on the two second gradient images to obtain the class structure tensor image.

[0007] Preferably, the convolution operation on the grayscale image based on the Sobel operator includes: The convolution kernel of the sobel operator is patched from a 3x3 matrix to a 5x5 matrix to obtain the IVE_sobel_like operator; A convolution operation is performed on the grayscale image through the IVE_sobel_like operator.

[0008] Preferably, the convolution calculation formulas of the first gradient images in the horizontal direction and the vertical direction are as follows: in, is the pixel value of the grayscale image, is the pixel value of the first gradient image in the horizontal direction, is the pixel value of the first gradient image in the vertical direction, is the convolution kernel vector of the IVE_sobel_like operator.

[0009] Preferably, the grayscale processing of the regional image to obtain a grayscale image includes: Gray-converting the 8-bit region image into the 8-bit grayscale image by combining two IVE_add operators; The grayscale formula is as follows: in, is the pixel value of the grayscale image, , , They are the pixel values ​​corresponding to the three primary colors of red, green and blue respectively.

[0010] Preferably, the convolution operation is performed on the grayscale image based on the Sobel operator to obtain first gradient images in the horizontal direction and the vertical direction respectively, including: Performing a convolution operation on the 8-bit grayscale image by using the IVE_sobel_like operator to obtain 16-bit first gradient images in the horizontal direction and the vertical direction respectively; The step of quantizing and converting the two first gradient images into absolute values ​​to obtain a second gradient image includes: The two 16-bit first gradient images are quantized into 8-bit gradient images respectively through two IVE_16bit_to_8bit operators, and absolute values ​​are performed at the same time to obtain the second gradient image.

[0011] Preferably, the binarization processing of the class structure tensor image to obtain a binarized image of the class structure tensor image includes: Performing a binarization operation of a threshold value of T [0, 255] on the class structure tensor image through the IVE_threshold operator to obtain a binarized image of the class structure tensor image, T∈(0, 255); The step of performing quantitative statistics on the maximum values ​​according to the binary image of the class structure tensor image to calculate the texture complexity includes: The IVE_hist operator is used to perform a number count of the maximum value 255 on the binary image of the class structure tensor image to obtain the texture information amount; The texture complexity is calculated according to the proportion of the texture information in the total pixels; Among them, the number of maximum values ​​255 is counted by the following formula: in, is the pixel value of the binary image of the class structure tensor image, For statistical results, Express judgment Is it true? If so, mark it as 1, otherwise mark it as 0.

[0012] According to the second aspect of the present application, a device for calculating regional image texture complexity based on IVE is proposed, comprising: An acquisition module is configured to acquire an RGB image to be processed, determine a region of interest according to the RGB image, and acquire coordinate information of the region of interest; A transport module is configured to transport and save the region of interest using a DMA operator of an IVE hardware acceleration module according to the coordinate information of the region of interest to obtain a region image; A processing module configured to calculate the structure tensor of the regional image through the image processing operator in the IVE hardware acceleration module to obtain a quasi-structure tensor image; The computing module is configured to perform binarization processing on the class structure tensor image to obtain a binarized image of the class structure tensor image; perform quantitative statistics on the maximum value according to the binarized image of the class structure tensor image to calculate the texture complexity.

[0013] According to the third aspect of the present application, an electronic device is proposed, comprising: one or more processors; a memory for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the electronic device implements the regional image texture complexity calculation method based on IVE as provided in any embodiment of the first aspect above.

[0014] According to the fourth aspect of the present application, a computer-readable storage medium is proposed, on which a computer program is stored. When the program is executed by a processor, the method for calculating regional image texture complexity based on IVE is implemented as provided in any embodiment of the first aspect above.

[0015] The present application proposes a method, device and storage medium for calculating the texture complexity of a regional image based on IVE. The DMA operator in the IVE hardware acceleration module is used to move and save the region of interest, and then the structure tensor of the regional image is calculated by combining the image processing operators in the IVE hardware acceleration module. Compared with the traditional method of calculating the structure tensor of the regional image based on the image processing library such as OPENCV on the CPU, the present application greatly reduces the amount of calculation while ensuring the performance and effect, and realizes the acceleration of the calculation of the structure tensor of the regional image, thereby realizing the acceleration of the calculation of the texture complexity of the regional image. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The accompanying drawings are included to provide a further understanding of the embodiments and are incorporated into and constitute a part of this specification. The accompanying drawings illustrate the embodiments and are used together with the description to explain the principles of the present invention. It will be easy to recognize other embodiments and many expected advantages of the embodiments because they become better understood by reference to the following detailed description. The elements of the drawings are not necessarily to scale with each other. The same reference numerals refer to corresponding similar parts.

[0017] Figure 1 is a flow chart of a method for calculating regional image texture complexity based on IVE according to a specific embodiment of the present application; Figure 2 is an architecture diagram of an IVE hardware acceleration module for calculating regional image texture complexity according to a specific embodiment of the present application; Figure 3 is a flowchart of obtaining a class structure tensor image according to a specific embodiment of the present application; Figure 4 is a structural diagram of a regional image texture complexity calculation device based on IVE according to a specific embodiment of the present application; Figure 5 is a schematic diagram of an electronic device according to a specific embodiment of the present application. DETAILED DESCRIPTION

[0018] The features and exemplary embodiments of various aspects of the present invention will be described in detail below. In order to make the purpose, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below in conjunction with the accompanying drawings and Examples. It should be understood that the specific embodiments described herein are only configured to explain the present invention and are not configured to limit the present invention. For those skilled in the art, the present invention can be implemented without the need for some of these specific details. The following description of the embodiments is only to provide a better understanding of the present invention by illustrating examples of the present invention.

[0019] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the statement "include..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0020] This application proposes a method for calculating regional image texture complexity based on IVE. Figure 1 FIG. 4 shows a flow chart of a method for calculating regional image texture complexity based on IVE according to a specific embodiment of the present application. Figure 1 As shown, the method comprises the following steps: Step S10: obtaining an RGB image to be processed, determining a region of interest according to the RGB image, and obtaining coordinate information of the region of interest.

[0021] Specifically, an RGB image containing the target is captured by an image acquisition device, the first frame of the image is selected as the RGB image to be processed, and then the coordinate information of the region of interest is sent by the upper computer according to the ROI region of interest manually selected by the user, and the lower computer receives the coordinate information.

[0022] In this embodiment, the image acquisition device is a drone gimbal camera integrated with an AI module.

[0023] Step S20: According to the coordinate information of the region of interest, the DMA operator of the IVE hardware acceleration module is used to move and save the region of interest to obtain a region image.

[0024] Step S30: Calculate the structure tensor of the regional image through the image processing operator in the IVE hardware acceleration module to obtain a quasi-structure tensor image.

[0025] Figure 2 FIG. 4 shows an architecture diagram of an IVE hardware acceleration module for calculating regional image texture complexity according to a specific embodiment of the present application. Figure 3 FIG. 4 shows a flowchart of obtaining a class structure tensor image according to a specific embodiment of the present application, as shown in FIG. Figure 2 , Figure 3 As shown, step S30 specifically includes: Step S301: grayscale the region image to obtain a grayscale image.

[0026] In a specific embodiment, two IVE_add operators are combined to convert an 8-bit regional image into an 8-bit grayscale image. Specifically, an IVE_add operator is used to perform a superposition calculation on two primary colors of the RGB three primary colors, and then another IVE_add operator is used to perform a second superposition calculation on the result of the superposition calculation and the remaining primary color to obtain an 8-bit grayscale image. The grayscale conversion formula is as follows: in, is the pixel value of the grayscale image, , , They are the pixel values ​​corresponding to the three primary colors of red, green and blue respectively.

[0027] Step S302: performing a convolution operation on the grayscale image based on the Sobel operator to obtain first gradient images in the horizontal direction and the vertical direction respectively.

[0028] In a specific embodiment, since the IVE hardware acceleration module can only adapt to the convolution filter template of the 5x5 matrix, it is necessary to first patch the convolution kernel of the traditional sobel operator from a 3x3 matrix to a 5x5 matrix to obtain the IVE_sobel_like operator.

[0029] The specific patch operations are as follows:

[0030] Among them, the matrix is the convolution kernel of the Sobel operator (before and after transposition), the matrix It is the convolution kernel of IVE_sobel_like operator (before and after transposition).

[0031] After the patch obtains the IVE_sobel_like operator, the 8-bit grayscale image is convolved once by the IVE_sobel_like operator to obtain a 16-bit first vertical gradient image, and then the IVE_sobel_like operator is transposed and convolved again to obtain a 16-bit first horizontal gradient image. The convolution calculation formulas for the first gradient images in the horizontal and vertical directions are as follows: in, is the pixel value of the grayscale image, is the pixel value of the first gradient image in the horizontal direction, is the pixel value of the first gradient image in the vertical direction, It is the convolution kernel vector of IVE_sobel_like operator.

[0032] Step S303: quantizing and converting the two first gradient images into absolute values ​​to obtain a second gradient image.

[0033] In a specific embodiment, two 16-bit first gradient images are quantized into 8-bit gradient images through two IVE_16bit_to_8bit operators respectively to reduce the amount of calculation, and the 8-bit gradient images are converted to absolute values ​​to obtain the second gradient images.

[0034] Step S304: performing weighted superposition calculation on the two second gradient images to obtain a class structure tensor image.

[0035] In a specific embodiment, a weighted superposition calculation of coefficients (0.5, 0.5) is performed on two second gradient images through an IVE_add operator to obtain a structure-like tensor image.

[0036] It should be noted that, in other implementations, different weight coefficients may be used to perform weighted superposition calculation on the two second gradient images according to actual needs, which is not limited here.

[0037] In short, step S303 and step S304 are equivalent to performing the following calculations on the first gradient image in the horizontal direction and the vertical direction: in, The pixel values ​​of the class structured tensor image.

[0038] In the traditional method, the regional image structure tensor is calculated on the CPU based on image processing libraries such as OPEN CV, usually by performing the following calculations on the first gradient images in the horizontal and vertical directions: On the one hand, the traditional method requires more complex mathematical calculations than the method of the present application, and is slow and inefficient in CPU-based image operations, and the efficiency decreases as the image size increases; on the other hand, no image processing operator matching the traditional method can be found on the IVE hardware acceleration module.

[0039] Therefore, an important inventive point of the present application is that a simpler mathematical calculation is used to calculate the regional image structure tensor. At the same time, the IVE_16bit_to_8bit operator and IVE_add operator adapted to the mathematical calculation are found in the IVE hardware acceleration module. The two operators are used to realize quantization, absolute value conversion and weighted superposition calculation, thereby realizing the acceleration of the calculation of the structure tensor.

[0040] After step S30, the method further includes: Step S40: binarizing the class structure tensor image to obtain a binarized image of the class structure tensor image.

[0041] In a specific embodiment, the class structure tensor image is binarized with a threshold value of T [0, 255] using the IVE_threshold operator.

[0042] In this embodiment, T is set to 128, that is, pixel values ​​less than 128 are all classified as 0, and pixel values ​​greater than 128 are all classified as 255, thereby obtaining a binary image of a structured tensor image. In other embodiments, T can be selected as an empirical value, which is not limited here.

[0043] Step S50: Count the maximum values ​​according to the binary image of the class structure tensor image to calculate the texture complexity.

[0044] In a specific embodiment, the IVE_hist operator is used to perform quantitative statistics of the maximum value 255 on the binary image of the class structure tensor image to obtain the texture information amount, and then the texture complexity is calculated according to the proportion of the texture information amount in the total pixels (height x width).

[0045] Specifically, the number of maximum values ​​255 is counted using the following formula: in, is the pixel value of the binary image of the class structure tensor image, For statistical results, Express judgment Is it true? If so, mark it as 1, otherwise mark it as 0.

[0046] It should be noted that, although only the maximum value 255 needs to be counted in this embodiment, the IVE_hist operator does not only count 255, but traverses and counts from 0 to 255, which will lose some performance to a certain extent, but for simple quantity statistics, the overall impact on performance is not significant.

[0047] Those skilled in the art will appreciate that, in other implementations, only the maximum value 255 may be counted by other calculation methods, and the calculation method may be to reference an external module or to find an operator supporting the calculation method from the IVE hardware acceleration module.

[0048] In summary, several important invention points of this application are: 1. In the process of calculating the regional image structure tensor for texture complexity calculation, the IVE hardware acceleration module is introduced to achieve accelerated calculation while ensuring performance and effect.

[0049] 2. The calculation of the regional image structure tensor adopts simpler mathematical calculations, and can be calculated by matching corresponding operators in the IVE hardware acceleration module. Compared with the traditional method, the amount of calculation is greatly reduced, and the calculation of the structure tensor is accelerated.

[0050] 3. The deployment of the overall architecture of the IVE hardware acceleration module realizes the calculation of texture complexity through simple and efficient IVE image processing operators, reduces the overall calculation amount, and improves performance.

[0051] The present application proposes a method for calculating the texture complexity of a regional image based on IVE, which uses the DMA operator in the IVE hardware acceleration module to move and save the region of interest, and then calculates the structure tensor of the regional image through a combination of image processing operators in the IVE hardware acceleration module. Compared with the traditional method of calculating the structure tensor of the regional image based on image processing libraries such as OPEN CV on the CPU, the present application greatly reduces the amount of calculation while ensuring performance and effect, and realizes the acceleration of the calculation of the structure tensor of the regional image, thereby realizing the acceleration of the calculation of the texture complexity of the regional image.

[0052] According to the above-mentioned method for calculating regional image texture complexity based on IVE, based on the same inventive concept, the present application also proposes a device for calculating regional image texture complexity based on IVE. Figure 4 FIG. 4 shows a structural diagram of a regional image texture complexity calculation device based on IVE according to a specific embodiment of the present application. Figure 4 As shown, the device comprises: The acquisition module 100 is configured to acquire the RGB image to be processed, determine the region of interest according to the RGB image, and acquire the coordinate information of the region of interest; The transport module 200 is configured to transport and save the region of interest using the DMA operator of the IVE hardware acceleration module according to the coordinate information of the region of interest to obtain a region image; The processing module 300 is configured to calculate the structure tensor of the regional image through the image processing operator in the IVE hardware acceleration module to obtain a quasi-structure tensor image; The calculation module 400 is configured to perform binarization processing on the class structure tensor image to obtain a binarized image of the class structure tensor image; perform quantitative statistics on the maximum values ​​according to the binarized image of the class structure tensor image to calculate the texture complexity.

[0053] According to the above-mentioned IVE-based regional image texture complexity calculation method, based on the same inventive concept, the present application also proposes an electronic device. Figure 5 A schematic diagram of an electronic device according to a specific embodiment of the present application is shown. Figure 5 As shown, the electronic device includes: one or more processors 501, a memory 502, a bus 503 and a communication interface 504. Among them, the one or more processors 501, the memory 502 and the communication interface 504 are connected via the bus 503. The memory 502 is used to store one or more programs. When the one or more programs are executed by the one or more processors 501, the electronic device implements the regional image texture complexity calculation method based on IVE provided in any of the above embodiments.

[0054] According to the above-mentioned method for calculating regional image texture complexity based on IVE, based on the same inventive concept, the present application also proposes a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the method for calculating regional image texture complexity based on IVE provided in any of the above-mentioned embodiments is implemented.

[0055] In the embodiments of the present application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device / system / method embodiments described above are only schematic. For example, the division of the units can be a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0056] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0057] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0058] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk and other media that can store program codes.

[0059] Obviously, those skilled in the art can make various modifications and changes to the embodiments of the present invention without departing from the spirit and scope of the present invention. In this way, if these modifications and changes are within the scope of the claims of the present invention and their equivalents, the present invention is also intended to cover these modifications and changes. The word "comprising" does not exclude the presence of other elements or steps not listed in the claims. The simple fact that certain measures are recorded in mutually different dependent claims does not indicate that the combination of these measures cannot be used to profit. Any reference numerals in the claims should not be considered to limit the scope.

Claims

1. A method for calculating regional image texture complexity based on IVE, characterized in that: include: Acquire an RGB image to be processed, determine a region of interest according to the RGB image, and acquire coordinate information of the region of interest; According to the coordinate information of the region of interest, the region of interest is moved and saved using the DMA operator of the IVE hardware acceleration module to obtain a regional image; Calculating the structure tensor of the regional image by using the image processing operator in the IVE hardware acceleration module to obtain a quasi-structure tensor image; Binarizing the class structure tensor image to obtain a binary image of the class structure tensor image; According to the binarized image of the class structure tensor image, the maximum value is quantitatively counted to calculate the texture complexity.

2. The method according to claim 1, characterized in that: The step of calculating the structure tensor of the regional image by using the image processing operator in the IVE hardware acceleration module to obtain a quasi-structure tensor image includes: Performing grayscale processing on the regional image to obtain a grayscale image; Performing a convolution operation on the grayscale image based on a Sobel operator to obtain first gradient images in horizontal and vertical directions respectively; quantizing and converting the two first gradient images into absolute values ​​to obtain a second gradient image; A weighted superposition calculation is performed on the two second gradient images to obtain the class structure tensor image.

3. The method according to claim 2, characterized in that The convolution operation on the grayscale image based on the Sobel operator includes: The convolution kernel of the sobel operator is patched from a 3x3 matrix to a 5x5 matrix to obtain the IVE_sobel_like operator; A convolution operation is performed on the grayscale image through the IVE_sobel_like operator.

4. The method according to claim 3, characterized in that The convolution calculation formulas of the first gradient image in the horizontal direction and the vertical direction are as follows: in, is the pixel value of the grayscale image, is the pixel value of the first gradient image in the horizontal direction, is the pixel value of the first gradient image in the vertical direction, is the convolution kernel vector of the IVE_sobel_like operator.

5. The method according to claim 2, characterized in that: The grayscale processing of the region image to obtain a grayscale image includes: Gray-converting the 8-bit region image into the 8-bit grayscale image by combining two IVE_add operators; The grayscale formula is as follows: in, is the pixel value of the grayscale image, , , They are the pixel values ​​corresponding to the three primary colors of red, green and blue respectively.

6. The method according to claim 3, characterized in that The convolution operation is performed on the grayscale image based on the Sobel operator to obtain first gradient images in the horizontal direction and the vertical direction respectively, including: Performing a convolution operation on the 8-bit grayscale image by using the IVE_sobel_like operator to obtain 16-bit first gradient images in the horizontal direction and the vertical direction respectively; The step of quantizing and converting the two first gradient images into absolute values ​​to obtain a second gradient image includes: The two 16-bit first gradient images are quantized into 8-bit gradient images respectively through two IVE_16bit_to_8bit operators, and absolute values ​​are performed at the same time to obtain the second gradient image.

7. The method according to claim 1, characterized in that The binarization process is performed on the class structure tensor image to obtain a binarized image of the class structure tensor image, including: Performing a binarization operation of a threshold value of T [0, 255] on the class structure tensor image through the IVE_threshold operator to obtain a binarized image of the class structure tensor image, T∈(0, 255); The step of performing quantitative statistics on the maximum values ​​according to the binary image of the class structure tensor image to calculate the texture complexity includes: The IVE_hist operator is used to perform a number count of the maximum value 255 on the binary image of the class structure tensor image to obtain the texture information amount; The texture complexity is calculated according to the proportion of the texture information in the total pixels; Among them, the number of maximum values ​​255 is counted by the following formula: in, is the pixel value of the binary image of the class structure tensor image, For statistical results, Express judgment Is it true? If so, mark it as 1, otherwise mark it as 0.

8. A device for calculating regional image texture complexity based on IVE, characterized in that: include: An acquisition module is configured to acquire an RGB image to be processed, determine a region of interest according to the RGB image, and acquire coordinate information of the region of interest; A transport module is configured to transport and save the region of interest using a DMA operator of an IVE hardware acceleration module according to the coordinate information of the region of interest to obtain a region image; A processing module configured to calculate the structure tensor of the regional image through the image processing operator in the IVE hardware acceleration module to obtain a quasi-structure tensor image; The computing module is configured to perform binarization processing on the class structure tensor image to obtain a binarized image of the class structure tensor image; perform quantitative statistics on the maximum value according to the binarized image of the class structure tensor image to calculate the texture complexity.

9. An electronic device, characterized in that: include: one or more processors; A memory for storing one or more programs, which, when executed by the one or more processors, enables the electronic device to implement the method as claimed in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images