A method, device, electronic device and medium for calculating discovery probability

Abstract

This application relates to the field of image processing technology, and specifically provides a method, device, equipment, and medium for calculating discovery probability. The method for calculating discovery probability in this application includes: based on a preset target box, determining a target area and a background area from the image to be detected, extracting a first feature vector corresponding to the target area and a second feature vector corresponding to the background area, calculating the feature similarity between the target area and the background area based on the first feature vector and the second feature vector, and calculating the discovery probability of a preset detection target in the image to be detected based on the feature similarity. Through the above method, the accuracy of calculating the discovery probability is effectively improved, the problem of timely determining whether the detection target is discovered is solved, and the ability to formulate countermeasures in advance is improved.

Description

Technical Field

[0001] The present application relates to the technical field of image processing, and in particular, to a discovery probability calculation method, device, electronic device, and computer-readable storage medium. Background Art

[0002] With the development of camouflage technology, the recognizable information of battlefield camouflage targets (such as camouflaged tanks, camouflaged combat vehicles, and tactical squads) is decreasing, but it is particularly important to quickly identify them under actual combat conditions (especially in complex battlefield environments such as smoke, fog, dust, and haze).

[0003] Currently, there are many methods for target detection in battlefield environments. Traditional target detection technologies include: radar detection, infrared image recognition, visible light images, spectral imaging recognition, etc. However, the target recognition technology based on radar detection needs to emit electromagnetic waves to the target, which is easily affected by the complex electromagnetic environment of the battlefield and there is a possibility of being detected by the enemy; although the infrared image recognition technology has strong target monitoring capabilities, with the rapid development of infrared stealth technology, it restricts the application of this technology to a certain extent; visible light images have the ability to recognize the fine morphological images of the battlefield environment, but they are powerless to detect camouflaged targets; when using spectral imaging means for detection, for targets with spectral camouflage, the phenomena of "same spectrum, different objects" and "same object, different spectra" may occur, and there will be a certain false alarm rate in complex target detection. For the above reasons, the existing target recognition technologies have the defect of low accuracy in the recognition results of camouflaged targets. Summary of the Invention

[0004] To solve the above problems, the present application provides a discovery probability calculation method, including: based on a preset target box, determining a target area and a background area from the image to be detected, extracting a first feature vector corresponding to the target area and a second feature vector corresponding to the background area, calculating the feature similarity between the target area and the background area based on the first feature vector and the second feature vector, and calculating the discovery probability of a preset detection target in the image to be detected based on the feature similarity. Through the above detection method, the accuracy of calculating the discovery probability can be effectively improved, the problem of timely determining whether the detection target is detected can be solved, and the ability to formulate countermeasures in advance can be improved.

[0005] In a first aspect, an embodiment of the present application provides a discovery probability calculation method, including: based on a preset target box, determining a target area and a background area from the image to be detected; extracting a first feature vector corresponding to the target area and a second feature vector corresponding to the background area; calculating the feature similarity between the target area and the background area based on the first feature vector and the second feature vector; and calculating the discovery probability of a preset detection target in the image to be detected based on the feature similarity.

[0006] In a second aspect, an embodiment of the present application provides a discovery probability calculation device, including: an acquisition module, configured to determine a target area and a background area from an image to be detected based on a preset target box; an extraction module, configured to extract a first feature vector of the target area and a second feature vector of the background area; a first calculation module, configured to calculate a feature similarity between the target area and the background area based on the first feature vector and the second feature vector; and a second calculation module, configured to calculate a discovery probability of a preset detection target in the image to be detected based on the feature similarity.

[0007] In a third aspect, an embodiment of the present application provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the method described in the first aspect above is implemented.

[0008] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the method described in the first aspect above.

[0009] The technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages:

[0010] By determining a target area and a background area from an image to be detected based on a preset target box, and extracting a first feature vector of the target area and a second feature vector of the background area, the embodiments of the present application can improve the image processing ability for complex images to be detected and improve the accuracy of calculating the discovery probability of a preset detection target in the subsequent image to be detected.

[0011] Furthermore, by calculating the feature similarity between the target area and the background area based on the first feature vector and the second feature vector, and calculating the discovery probability of a preset detection target in the image to be detected based on the feature similarity, the discovery probability of the detection target can be accurately obtained, so as to accurately determine whether the detection target is discovered, effectively improve the accuracy of image detection, enhance the timeliness of determining whether the detection target is discovered, enhance the accuracy of camouflage effect evaluation, and thus enhance the ability to formulate countermeasures in advance.

[0012] The additional aspects and advantages of the present application will be partially given in the following description, partially become obvious from the following description, or be understood through the practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Various other advantages and benefits will become clear to those of ordinary skill in the art by reading the following detailed description of the preferred embodiments. The drawings are only for the purpose of showing the preferred embodiments and are not considered to be a limitation of the present application. Also, throughout the drawings, the same reference numerals are used to denote the same components. In the drawings:

[0014] Figure 1 A flowchart of a discovery probability calculation method provided by an embodiment of the present application is shown;

[0015] Figure 2 A schematic structural diagram of a feature extraction network provided by an embodiment of the present application is shown;

[0016] Figure 3 A curve graph of the fitting result of the mapping function from feature similarity to discovery probability provided by an embodiment of the present application is shown;

[0017] Figure 4 A schematic structural diagram of a discovery probability calculation device provided by an embodiment of the present application is shown;

[0018] Figure 5 A schematic structural diagram of an electronic device provided by an embodiment of the present application is shown. Detailed Embodiments

[0019] The exemplary embodiments of the present application will be described in more detail below with reference to the drawings. Although the exemplary embodiments of the present application are shown in the drawings, it should be understood that the present application can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided so that the present application can be more thoroughly understood and the scope of the present application can be fully conveyed to those skilled in the art.

[0020] With the development of camouflage technology, the recognizable information of battlefield camouflage targets (such as camouflaged tanks, camouflaged combat vehicles, and tactical squads) is decreasing, but it is particularly important to quickly identify them under actual combat conditions (especially in complex battlefield environments such as smoke, fog, dust, and haze). Currently, there are many methods for target detection in battlefield environments. Traditional target detection technologies include: radar detection, infrared image recognition, visible light images, spectral imaging recognition, etc. However, the target recognition technology based on radar detection needs to emit electromagnetic waves to the target, which is easily affected by the complex electromagnetic environment of the battlefield and there is a possibility of being detected by the enemy; although the infrared image recognition technology has strong target monitoring capabilities, with the rapid development of infrared stealth technology, it restricts the application of this technology to a certain extent; visible light images have the ability to recognize the fine morphological images of the battlefield environment, but they are powerless in detecting camouflaged targets; when using spectral imaging means for detection, for targets with spectral camouflage, the phenomena of "same spectrum, different objects" and "same object, different spectra" may occur, and there will be a certain false alarm rate in detecting complex targets. For the above reasons, the existing target recognition technologies have the defect of low accuracy in the recognition results of camouflaged targets.

[0021] Based on this, the embodiments of the present application provide a discovery probability calculation method. The following specifically describes the solution of the embodiments of the present application with reference to the drawings.

[0022] See Figure 1 The flowchart of a discovery probability calculation method shown, the method specifically includes the following steps:

[0023] Step 101: Based on a preset target box, determine the target area and the background area from the image to be detected.

[0024] In the embodiments of the present application, the preset target box refers to the target box used to determine the target area and the background area in the image to be detected. The preset target box can be a target box calculated by those skilled in the art based on experimental data, or a target box obtained by those skilled in the art after adjusting the already set target box according to actual needs. The embodiments of the present application do not make specific limitations.

[0025] In one implementation manner, the input image can be detected by a trained image detection model to obtain the features of the image to be detected, and then based on the preset target box, the target area and the background area features are determined from the image to be detected.

[0026] Specifically, the image detection model may include a feature extraction network and a feature fusion network. The input image may include a first color image and a first grayscale image, or may be any image. The first color image refers to an image taken under sufficient visible light, such as an image taken during the day, when there are a certain number of lights, etc.; the first grayscale image may be an image taken under insufficient visible light or no visible light, or may be an image taken by an infrared device. The embodiments of the present application do not make specific limitations.

[0027] Further, if the input image is the first color image and the first grayscale image, then the first color image and the first grayscale image are input into the feature extraction network, and the first feature is output by the feature extraction network. Specifically, the feature extraction network may include a first feature extraction module, a second feature extraction module, a first feature fusion module, and a feature extraction module. Refer to Figure 2 As shown, the first complementary feature and the first redundant feature of the first color image can be extracted by the first feature extraction module, and the first complementary feature and the first redundant feature are transmitted to the first feature fusion module. The second complementary feature and the second redundant feature of the first grayscale image are extracted by the second feature extraction module, and the second complementary feature and the second redundant feature are transmitted to the first feature fusion module. Then, the first complementary feature and the second complementary feature are fused by the first feature fusion module to obtain a third complementary feature, the first redundant feature and the second redundant feature are fused to obtain a third redundant feature, the third complementary feature and the third redundant feature are combined to obtain the combined feature of the image to be detected and transmitted to the feature extraction module. Finally, the feature extraction module performs feature extraction processing on the combined feature to obtain the first feature.

[0028] Further, the feature extraction network has a total of five layers, and each layer will output a feature map information after downsampling processing. The downsampling multiple can be preset by those skilled in the art according to experimental data, or can be obtained by adjusting the preset multiple according to actual needs. The embodiments of the present application do not make specific limitations. Preferably, the downsampling multiple adopted in the embodiments of the present application is 2 times.

[0029] Further, the first feature extraction module includes a first layer and a second layer. A max pooling layer is provided between the first layer and the second layer. The first layer includes a 1×1 convolutional layer 1, a normalization layer 1, and a RELU activation function connected in sequence. The second layer includes a 1×1 convolutional layer 2, a normalization layer 2, a RELU activation function, a 1×1 convolutional layer 3, a normalization layer 3, a RELU activation function, a 1×1 convolutional layer 4, a normalization layer 4, and a RELU activation function connected in sequence. The second feature extraction module has the same structure as the first feature extraction module, and will not be repeated here.

[0030] By setting two layers for the first feature extraction module and the second feature extraction module, useful image features in the image can be effectively extracted, ensuring the alignment ability of pixels in the image features, improving the feature extraction ability of the image. At the same time, using two layers can also reduce the computational amount and improve the running speed.

[0031] Furthermore, the first feature is transmitted to the feature fusion network to obtain the second feature. Specifically, the feature extraction module includes a third level, a fourth level, and a fifth level, and the structures of the third level, the fourth level, and the fifth level are the same as that of the second level, which will not be repeated here.

[0032] It should be noted that complementary features refer to the color feature information of the target image contained in the first color image and the gray feature information of the target image contained in the first grayscale image.

[0033] Redundant features refer to the color feature information of other images except the target image contained in the first color image and the gray feature information of other images except the target image contained in the first grayscale image.

[0034] Furthermore, if the input image is an arbitrary image, the trained image detection model can directly detect the input image.

[0035] Furthermore, the preset target box can include the horizontal and vertical coordinates, length, width, and rotation angle of the center point of the target box. Based on the target detection algorithm Mask R-CNN, the target region and the background region can be determined according to the preset target box. The Mask R-CNN algorithm is a deep learning-based target detection method, inherited from Faster R-CNN. Mask R-CNN adds a Mask Prediction Branch on top of Faster R-CNN and proposes ROI Align based on ROIPooling.

[0036] Furthermore, based on the horizontal and vertical coordinates and the rotation angle of the center point of the target box, the position of the target box in the image to be detected is determined. The image region covered by the target box within the image to be detected is used as the target region. The length and width of the target box are respectively increased by a preset value to obtain a new target box. The image region covered by the new target box within the image to be detected and not including the target region is used as the background region.

[0037] Step 102: Extract the first feature vector corresponding to the target region and the second feature vector corresponding to the background region.

[0038] In the embodiments of the present application, the first feature vector may include a first threshold number of first grid feature vectors, and the second feature vector may include a second threshold number of second grid feature vectors.

[0039] In one implementation, based on the length and width of the target box, the target area may be divided into a first threshold number of first grids. For each first grid, a preset first number of first candidate sampling points are obtained within the first grid. Bilinear interpolation processing is respectively performed on each first candidate sampling point to obtain the first candidate sampling features corresponding to each first candidate sampling point. An average value calculation is performed on all the first candidate sampling features to obtain the first grid feature vector corresponding to the first grid. The set of the first threshold number of first grid feature vectors is used as the first feature vector corresponding to the target area.

[0040] Further, based on the length and width of the new target box, the background area may be divided into a second threshold number of second grids. For each second grid, a preset second number of second candidate sampling points are obtained within the second grid. Bilinear interpolation processing is respectively performed on each second candidate sampling point to obtain the second candidate sampling features corresponding to each second candidate sampling point. An average value calculation is performed on all the second candidate sampling features to obtain the second grid feature vector corresponding to the second grid. The set of the second threshold number of second grid feature vectors is used as the second feature vector corresponding to the background area.

[0041] Step 104: Calculate the feature similarity between the target area and the background area based on the first feature vector and the second feature vector.

[0042] In one implementation, the first feature vector and the second feature vector may be aligned, and then based on the aligned first feature vector and the second feature vector, the feature similarity between the target area and the background area is calculated.

[0043] Further, the target area may include a plurality of first grids, the background area may include a plurality of second grids, the first feature vector may include the first grid feature vectors corresponding to each first grid, and the second feature vector may include the second grid feature vectors corresponding to each second grid. The alignment processing of the first feature vector and the second feature vector may specifically be, for each second grid among the plurality of second grids: obtaining the first grid closest to the position of the second grid as the candidate first grid, and then corresponding the second grid feature vector corresponding to the second grid with the first grid feature vector corresponding to the candidate first grid.

[0044] Further, based on the first feature vector and the second feature vector after alignment processing, calculate the feature similarity between the target region and the background region. Specifically, it can be based on the first feature vector and the second feature vector after alignment processing, calculate the distance between the first feature vector and the second feature vector, and based on this distance, obtain the scaling factor of the image to be detected, and multiply the scaling factor by the distance to obtain the feature similarity between the target region and the background region.

[0045] Further, the length of the first feature vector or the second feature vector after alignment processing can be obtained, and based on the first feature vector, the second feature vector, and the length after alignment processing, the distance between the first feature vector and the second feature vector can be obtained through formula (1):

[0046]

[0047] where dist represents the distance between the first feature vector and the second feature vector, n represents the length, e a represents the first feature vector, and e b represents the second feature vector.

[0048] Further, obtaining the length of the first feature vector or the second feature vector after alignment processing can be: the preset channels of the image to be detected × the first feature vector × the second feature vector.

[0049] Further, before obtaining the scaling factor of the image to be detected, multiple images to be calculated can be prepared in advance, and the candidate distances between the first feature vector and the second feature vector corresponding to each image to be calculated can be obtained through the implementation manners of steps 101 to 103 above. The maximum value and the minimum value among the multiple candidate distances are obtained as the maximum distance and the minimum distance, and the scaling factor of the image to be detected is calculated through formula (2):

[0050]

[0051] where scale represents the scaling factor, dist min represents the maximum distance, and dist max represents the minimum distance.

[0052] It should be noted that if the distance between the first feature vector and the second feature vector is larger, the feature similarity is larger, indicating that the target region and the background region are less similar; if the distance between the first feature vector and the second feature vector is smaller, the feature similarity is smaller, indicating that the target region and the background region are more similar.

[0053] Step 104: Calculate the discovery probability of the preset detection target in the image to be detected based on the feature similarity.

[0054] In one embodiment, the discovery probability of a preset detection target in a to-be-detected image can be calculated by formula (3) based on the feature similarity and a preset third threshold value:

[0055]

[0056] where p(d) represents the discovery probability, k represents a constant, r represents the feature similarity, and q represents the preset third threshold value.

[0057] Further, the mapping function between the feature similarity and the discovery probability is fitted by the least squares method to obtain the constant k and the third threshold value. For the fitting result, see Figure 3 as shown.

[0058] In the embodiment of the present application, by determining the target region and the background region from the to-be-detected image based on the preset target box, and extracting the first feature vector of the target region and the second feature vector of the background region, the image processing ability for complex to-be-detected images can be improved, the loss of the target image can be avoided, the detection accuracy of the to-be-detected image can be improved, the computational amount of the algorithm can be reduced, the detection speed can be increased, and further the accuracy of calculating the discovery probability of the preset detection target in the subsequent to-be-detected image can be improved; in addition, by determining the target region and the background region, the features of different regions can be better compared and analyzed, the computational amount can be reduced, and the accuracy of calculating the discovery probability can also be improved.

[0059] Further, based on the first feature vector and the second feature vector, the feature similarity between the target region and the background region is calculated, and based on the feature similarity, the discovery probability of the preset detection target in the to-be-detected image is calculated, so that the discovery probability of the detection target can be accurately known, thereby accurately determining whether the detection target is discovered, effectively improving the accuracy of image detection, enhancing the timeliness of determining whether the detection target is discovered, and thus enhancing the ability to formulate countermeasures in advance.

[0060] See Figure 4 , the embodiment of the present application further provides a discovery probability calculation device, which is used to execute the discovery probability calculation method described in the above embodiment. The device includes:

[0061] An acquisition module 201, configured to acquire a target region and a background region from a to-be-detected image based on a preset target box;

[0062] An extraction module 202, configured to extract a first feature vector corresponding to the target region and a second feature vector corresponding to the background region;

[0063] A first calculation module 203, configured to calculate the feature similarity between the target region and the background region based on the first feature vector and the second feature vector;

[0064] A second computing module 204, configured to calculate a discovery probability of a preset detection target in an image to be detected based on feature similarity.

[0065] The image detection device provided by the embodiments of the present application and the discovery probability calculation method provided by the above embodiments are based on the same inventive concept and have the same beneficial effects as the methods adopted, run, or implemented by them.

[0066] The embodiments of the present application also provide an electronic device corresponding to the discovery probability calculation method provided by the foregoing embodiments. Please refer to Figure 5 , which shows a schematic diagram of an electronic device provided by some embodiments of the present application. As shown in the image detection, the electronic device 30 may include: a processor 300, a memory 301, a bus 302, and a communication interface 303. The processor 300, the communication interface 303, and the memory 301 are connected through the bus 302; a computer program that can run on the processor 300 is stored in the memory 301, and when the processor 300 runs the computer program, it executes the discovery probability calculation method provided by any of the foregoing embodiments of the present application.

[0067] Among them, the memory 301 may include a high-speed random access memory (RAM: Random Access Memory), and may also include a non-volatile memory, such as at least one disk memory. The communication connection between the system network element and at least one other network element is realized through at least one physical port 303 (which may be wired or wireless), and the Internet, wide area network, local area network, metropolitan area network, etc. can be used.

[0068] The bus 302 may be an ISA bus, a PCI bus, an EISA bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc. Among them, the memory 301 is used to store the program, and after receiving the execution instruction, the processor 300 executes the program. The discovery probability calculation method disclosed in any of the foregoing embodiments of the present application may be applied to the processor 300 or implemented by the processor 300.

[0069] The processor 300 may be an integrated circuit with the ability to process signals. In the implementation process, each step of the above method can be completed by the integrated logic circuit in the hardware of the processor 300 or instructions in the form of software. The above-mentioned processor 300 may be a general-purpose processor, including a central processing unit (CPU for short), a network processor (NP for short), etc.; it may also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. It can implement or execute the various methods, steps and logic block diagrams disclosed in the embodiments of the present application. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The steps of the method disclosed in combination with the embodiments of the present application can be directly embodied as being executed by the hardware decoding processor, or executed by a combination of the hardware and software modules in the decoding processor. The software module may be located in a mature storage medium in the art such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, or an electrically erasable programmable memory, a register, etc. This storage medium is located in the memory 301, and the processor 300 reads the information in the memory 301 and combines its hardware to complete the steps of the above method.

[0070] The electronic device provided by the embodiment of the present application and the discovery probability calculation method provided by the embodiment of the present application are based on the same inventive concept and have the same beneficial effects as the method adopted, run or implemented by it.

[0071] The embodiment of the present application also provides a computer-readable storage medium corresponding to the discovery probability calculation method provided by the foregoing embodiment, on which a computer program (i.e., a program product) is stored. When the computer program is run by a processor, it will execute the discovery probability calculation method provided by any of the foregoing embodiments.

[0072] It should be noted that examples of the computer-readable storage medium may also include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other optical and magnetic storage media, which will not be elaborated here one by one.

[0073] The embodiment of the present application also provides a computer program product corresponding to the discovery probability calculation method provided by the foregoing embodiment, including a computer program, and the computer program is executed by a processor to implement the discovery probability calculation method provided by the above embodiments.

[0074] The computer-readable storage medium and computer program product provided by the above embodiments of the present application are all based on the same inventive concept as the discovery probability calculation method provided by the embodiments of the present application, and have the same beneficial effects as the methods adopted, run, or implemented by the application programs stored therein.

[0075] It should be noted that:

[0076] The algorithms and displays provided herein are not inherently related to any particular computer, virtual device, or other equipment. Various general-purpose devices can also be used in conjunction with the teachings herein. The structure required to construct such devices will be apparent from the above description. In addition, the present application is not directed to any particular programming language. It should be understood that the content of the present application described herein can be implemented using various programming languages, and the description of a particular language above is to disclose the best mode of the present application.

[0077] In the specification provided herein, a large number of specific details are set forth. However, it can be understood that the embodiments of the present application can be practiced without these specific details. In some instances, well-known methods, structures, and technologies have not been shown in detail so as not to obscure the understanding of this specification.

[0078] Similarly, it should be understood that, in order to streamline the present application and assist in understanding one or more of the various inventive aspects, in the above description of the exemplary embodiments of the present application, the various features of the present application are sometimes grouped together into a single embodiment, figure, or description thereof. However, the disclosed method should not be construed as reflecting the intention that the claimed present application requires more features than are expressly recited in each claim. Rather, as reflected in the following claims, the inventive aspects lie in less than all the features of the single foregoing disclosed embodiment. Thus, the claims following the detailed description are hereby expressly incorporated into the detailed description, with each claim standing on its own as a separate embodiment of the present application.

[0079] Those skilled in the art can understand that the modules in the devices in the embodiments can be adaptively changed and arranged in one or more devices different from those of the embodiments. The modules or units or components in the embodiments can be combined into one module or unit or component, and in addition, they can be divided into multiple sub-modules or sub-units or sub-components. Except that at least some of such features and / or processes or units are mutually exclusive, any combination can be adopted to combine all the features disclosed in this specification (including the accompanying claims, abstract and drawings) and all the processes or units of any method or device so disclosed. Unless otherwise explicitly stated, each feature disclosed in this specification (including the accompanying claims, abstract and drawings) can be replaced by an alternative feature that provides the same, equivalent or similar purpose.

[0080] In addition, those skilled in the art can understand that although some of the embodiments described herein include certain features included in other embodiments rather than other features, the combination of the features of different embodiments means that it is within the scope of this application and forms different embodiments. For example, in the following claims, any one of the claimed embodiments can be used in any combination.

[0081] Each component embodiment of the present application can be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof. Those skilled in the art should understand that a microprocessor or a digital signal processor (DSP) can be used in practice to implement some or all of the functions of some or all of the components in the virtual machine creation device according to the embodiments of the present application. The present application can also be implemented as a device or device program (for example, a computer program and a computer program product) for executing part or all of the methods described herein. Such a program implementing the present application can be stored on a computer-readable medium, or can be in the form of one or more signals. Such signals can be downloaded from an Internet website, or provided on a carrier signal, or provided in any other form.

[0082] It should be noted that the above embodiments are illustrative of the present application rather than restrictive thereof, and those skilled in the art can design alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. The word "comprising" does not exclude the presence of elements or steps not listed in the claim. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The present application can be implemented by means of hardware including several different elements and by means of a suitably programmed computer. In the unit claims listing several devices, several of these devices can be embodied by the same item of hardware. The use of the words first, second, and third, etc. does not denote any order. These words can be interpreted as names.

[0083] As mentioned above, the above are only the preferred specific embodiments of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by those skilled in the art within the technical scope disclosed in the present application should be covered by the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the said claims.

Claims

1. A method for calculating discovery probability, characterized in that, Including: Determining a target region and a background region from an image to be detected based on a preset target box; The preset target box includes the abscissa and ordinate of the center point, length, height, and rotation angle; The determining the target region and the background region from the image to be detected based on the preset target box includes: determining the position of the target box in the image to be detected based on the abscissa and ordinate of the center point and the rotation angle, and taking the image region covered by the target box within the image to be detected as the target region; increasing the length and width of the target box by preset values respectively to obtain a new target box; taking the image region covered by the new target box within the image to be detected and not including the target region as the background region; Extracting a first feature vector corresponding to the target region and a second feature vector corresponding to the background region; the first feature vector includes a first threshold number of first grid feature vectors, and the second feature vector includes a second threshold number of second grid feature vectors; extracting the first feature vector corresponding to the target region and the second feature vector corresponding to the background region includes: dividing the target region into a first threshold number of first grids based on the length and width of the target box; for each of the first grids, obtaining a preset first number of first candidate sampling points within the first grid, respectively performing bilinear interpolation processing on each of the first candidate sampling points to obtain first candidate sampling features corresponding to each of the first candidate sampling points, calculating the average value of all the first candidate sampling features to obtain the first grid feature vector corresponding to the first grid; taking the set of the first threshold number of first grid feature vectors as the first feature vector corresponding to the target region; dividing the background region into a second threshold number of second grids based on the length and width of the new target box; for each of the second grids, obtaining a preset second number of second candidate sampling points within the second grid, respectively performing bilinear interpolation processing on each of the second candidate sampling points to obtain second candidate sampling features corresponding to each of the second candidate sampling points, calculating the average value of all the second candidate sampling features to obtain the second grid feature vector corresponding to the second grid; taking the set of the second threshold number of second grid feature vectors as the second feature vector corresponding to the background region; Calculating the feature similarity between the target region and the background region based on the first feature vector and the second feature vector; Calculating the discovery probability of a preset detection target in the image to be detected based on the feature similarity; 2. The method for calculating discovery probability according to claim 1, characterized in that, Calculating the feature similarity between the target region and the background region based on the first feature vector and the second feature vector includes: Performing alignment processing on the first feature vector and the second feature vector; Calculating the feature similarity between the target region and the background region based on the aligned first feature vector and the second feature vector.

3. The method for calculating discovery probability according to claim 2, characterized in that, The target area includes a plurality of first grids, the background area includes a plurality of second grids, the first feature vector includes a first grid feature vector corresponding to each first grid, and the second feature vector includes a second grid feature vector corresponding to each second grid. Aligning the first feature vector and the second feature vector includes: For each of the plurality of second grids: Obtain the first grid closest to the position of the second grid as the candidate first grid; Correspond the second grid feature vector corresponding to the second grid with the first grid feature vector corresponding to the candidate first grid.

4. The method for calculating discovery probability according to claim 2, characterized in that, Calculating the feature similarity between the target area and the background area based on the aligned first feature vector and second feature vector includes: Calculating the distance between the first feature vector and the second feature vector based on the aligned first feature vector and second feature vector; Obtaining a scaling factor of the image to be detected based on the distance; Multiplying the scaling factor by the distance to obtain the feature similarity between the target area and the background area.

5. The method for calculating discovery probability according to claim 1 or 3, characterized in that, Calculating the discovery probability of a preset detection target in the image to be detected based on the feature similarity includes: Calculating the discovery probability of a preset detection target in the image to be detected through formula (1) based on the feature similarity and a preset third threshold: Where p(d) represents the discovery probability, k represents a constant, r represents the feature similarity, and q represents a preset third threshold.

6. A discovery probability calculation device, adopting the method according to claim 1, characterized in that, Includes: An acquisition module, configured to determine a target area and a background area from the image to be detected based on a preset target box; An extraction module, configured to extract a first feature vector of the target area and a second feature vector of the background area; A first calculation module, configured to calculate the feature similarity between the target area and the background area based on the first feature vector and the second feature vector; A second calculation module, configured to calculate the discovery probability of a preset detection target in the image to be detected based on the feature similarity.

7. An electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, the method according to any one of claims 1-5 is implemented.

8. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, the method according to any one of claims 1-5 is implemented.

Images