A method and apparatus for fusion camouflage evaluation of overexposed images

Abstract

The application discloses a kind of overexposure image processing method and device for fusion camouflage evaluation, belong to image processing field.Method includes: respectively to the overexposure image obtained is carried out morphological filtering and AI identification processing, and sequentially obtain filter image and identification result;The filter image is carried out contour extraction, and according to extraction result and identification result sequentially determine the typical histogram distribution characteristics of target and original histogram distribution characteristics;According to the typical histogram distribution characteristics and the original histogram distribution characteristics, determine the similarity degree of target and corresponding typical target in overexposure image;According to the similarity degree, the pixel value of target in the overexposure image is handled, and clear image for doing fusion camouflage with other spectral images is obtained.The application can realize overexposure image and other spectral images are fused and carry out camouflage evaluation.

Description

Technical Field

[0001] This invention relates to the field of image processing technology, and in particular to an overexposed image processing method and apparatus for fusion camouflage assessment. Background Technology

[0002] Image fusion camouflage assessment technology is widely used in the military field. It measures camouflage methods such as camouflage nets and false targets using cameras in various spectral bands, and performs image feature extraction, feature fusion and quantitative analysis on the measurement results to obtain the camouflage effect of the camouflage methods. The visible spectrum is the most common spectrum carried by military satellites, drones and other reconnaissance equipment, and is widely used in the field of camouflage assessment.

[0003] In related technologies, common methods for camouflage assessment by fusing visible light with other spectral bands are based on visible light images with normal exposure times. However, in practical applications, there is a lack of methods for camouflage assessment using overexposed visible light images.

[0004] Therefore, there is an urgent need for an overexposed image processing method and apparatus for fusion camouflage assessment to solve the above-mentioned technical problems. Summary of the Invention

[0005] This invention provides a method and apparatus for fusing overexposed visible light images for camouflage assessment. It enables the fusion of overexposed visible light images with other spectral bands for camouflage assessment. The technical solution is as follows:

[0006] On the one hand, an overexposed image processing method for fusion camouflage assessment is provided, the method comprising:

[0007] The obtained overexposed images are subjected to morphological filtering and AI recognition processing respectively to obtain filtered images and recognition results.

[0008] Contour extraction is performed on the filtered image, and the typical histogram distribution features and original histogram distribution features of the target are determined sequentially based on the extraction results and recognition results;

[0009] Based on the typical histogram distribution characteristics and the original histogram distribution characteristics, the similarity between the target in the overexposed image and the corresponding typical target is determined;

[0010] The pixel values ​​of the target in the overexposed image are processed according to the similarity to obtain a clear image for camouflage fusion with other spectral band images.

[0011] On the other hand, an overexposed image processing apparatus for fusing camouflage assessment is provided, the apparatus comprising:

[0012] The first processing module is used to perform morphological filtering and AI recognition processing on the acquired overexposed image to obtain the filtered image and the recognition result in turn.

[0013] The first determining module is used to extract the contour of the filtered image and determine the typical histogram distribution features and the original histogram distribution features of the target in sequence based on the extraction results and the recognition results.

[0014] The second determining module is used to determine the similarity between the target in the overexposed image and the corresponding typical target based on the typical histogram distribution characteristics and the original histogram distribution characteristics.

[0015] The second processing module is used to process the pixel values ​​of the target in the overexposed image according to the similarity, so as to obtain a clear image for fusion and camouflage with other spectral band images.

[0016] On the other hand, a computer device is provided, the computer device including a memory and a processor, the memory for storing a computer program, and the processor for executing the computer program stored in the memory to implement the steps of the overexposed image processing method for fusion camouflage assessment described above.

[0017] On the other hand, a computer-readable storage medium is provided, wherein a computer program is stored therein, and when executed by a processor, the computer program implements the steps of the above-described overexposure image processing method for fusing camouflage assessment.

[0018] On the other hand, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps of the overexposed image processing method for fusing camouflage assessment described above.

[0019] The technical solution provided by this invention can bring at least the following beneficial effects: By performing morphological filtering and AI recognition processing on the acquired overexposed image, a filtered image and a recognition result are obtained sequentially; then, contour extraction is performed on the filtered image, and the typical histogram distribution features and the original histogram distribution features of the target are determined sequentially based on the extraction results and the recognition results; subsequently, the similarity between the target in the overexposed image and the corresponding typical target is determined based on the typical histogram distribution features and the original histogram distribution features; finally, the pixel values ​​of the target in the overexposed image are processed according to the similarity to obtain a clear image for fusion camouflage with other spectral band images. This method can realize target recognition and pixel replacement in overexposed visible light images under high confidence conditions. The target information in the replaced visible light image can be used for fusion camouflage evaluation with other spectral band images without affecting the accuracy of the fusion camouflage evaluation. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a flowchart of an overexposed image processing method for fusion camouflage assessment provided by an embodiment of the present invention;

[0022] Figure 2 This is a structural diagram of an overexposed image processing device for fusion camouflage assessment provided in an embodiment of the present invention;

[0023] Figure 3 This is a hardware architecture diagram of a computer device provided in an embodiment of the present invention. Detailed Implementation

[0024] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0025] As mentioned earlier, common methods for evaluating camouflage by fusing visible light with other spectral bands are based on visible light images with normal exposure times, and lack research on overexposed visible light images.

[0026] Based on this, the present invention provides an overexposed visible light image processing method, which can achieve fusion with other spectral band images and carry out camouflage assessment under the condition of overexposed visible light images.

[0027] The following describes the specific implementation of the above concept.

[0028] Please refer to Figure 1 This invention provides an overexposed image processing method for camouflage assessment, the method comprising:

[0029] Step 100: Perform morphological filtering and AI recognition processing on the obtained overexposed image to obtain the filtered image and recognition result in turn;

[0030] Step 102: Extract contours from the filtered image, and determine the typical histogram distribution features and the original histogram distribution features of the target in sequence based on the extraction results and recognition results;

[0031] Step 104: Based on the typical histogram distribution characteristics and the original histogram distribution characteristics, determine the similarity between the target in the overexposed image and the corresponding typical target;

[0032] Step 106: Process the pixel values ​​of the target in the overexposed image according to the similarity to obtain a clear image for fusion and camouflage with other spectral band images.

[0033] In this embodiment of the invention: Morphological filtering and AI recognition processing are applied to the acquired overexposed image to obtain a filtered image and a recognition result. Then, contour extraction is performed on the filtered image, and the typical histogram distribution features and original histogram distribution features of the target are determined sequentially based on the extraction and recognition results. Subsequently, the similarity between the target in the overexposed image and the corresponding typical target is determined based on the typical histogram distribution features and the original histogram distribution features. Finally, the pixel values ​​of the target in the overexposed image are processed according to the similarity to obtain a clear image for fusion camouflage with other spectral band images. This method can achieve target recognition and pixel replacement in overexposed visible light images under high confidence conditions. The target information in the replaced visible light image can be used for fusion camouflage evaluation with other spectral band images without affecting the accuracy of the fusion camouflage evaluation.

[0034] The following description Figure 1 The execution method of each step is shown.

[0035] First, for step 100, the obtained overexposed image is subjected to morphological filtering and AI recognition processing respectively, and the filtered image and recognition result are obtained in sequence.

[0036] Under overexposure conditions, visible light images lose some detail information, and the image accuracy cannot be guaranteed. Therefore, it is impossible to perform camouflage evaluation with other spectral images. In order to effectively solve the problem of missing detail information in visible light images, this embodiment of the invention uses the RK3588+FPGA method for processing.

[0037] Specifically, the FPGA chip receives the overexposed visible light image, performs a morphological filtering algorithm on the overexposed image, and sends the filtered image and the original overexposed image to the RK3588 chip through the image interface; the RK3588 chip receives the overexposed visible light image, runs an AI target recognition algorithm, and extracts the target location information, target location coordinates, and target classification results from the overexposed visible light image.

[0038] Then, for step 102, contour extraction is performed on the filtered image, and the typical histogram distribution features and original histogram distribution features of the target are determined sequentially based on the extraction results and recognition results.

[0039] In this embodiment of the invention, the contour extraction process includes: determining the minimum connected component coordinates of each target based on the edge contour information of all targets in the filtered image; and processing the number of image pixels occupied by each target based on the spatial resolution information to obtain the two-dimensional projected area of ​​each target.

[0040] Specifically, the RK3588 chip receives the filtered image, obtains the edge contour information of the target, obtains the minimum connected component of the target based on the edge contour information, counts the number of pixels occupied by each target in the image, and obtains the two-dimensional projected area of ​​each target by combining the spatial resolution information. The coordinates of the minimum connected component are then sent to the FPGA.

[0041] Further, based on the two-dimensional projected area and the recognition result, a matching is performed in the histogram distribution feature library to obtain the typical histogram distribution feature of the target's most likely corresponding category, as well as the number of the typical histogram distribution feature; target image data is extracted based on the coordinates of the minimum connected component, and histogram statistics are performed on the target image data to obtain the original histogram distribution feature of each target.

[0042] Specifically, the RK3588 matches the 2D projected area with the recognition result in a histogram distribution feature library to obtain the typical histogram distribution feature corresponding to the most likely classification, and sends the typical histogram distribution feature number to the FPGA. The typical histogram distribution feature is in 2D matrix form, suitable for storage in the FPGA's on-chip RAM. Pixel values ​​serve as FPGA address information, and the corresponding feature value is indexed through the FPGA address information. The FPGA receives the coordinates of the minimum connected component of the target, extracts the target image data based on the coordinates, and performs histogram statistics to obtain the histogram distribution features of each target.

[0043] For step 104, based on the typical histogram distribution characteristics and the original histogram distribution characteristics, the similarity between the target in the overexposed image and the corresponding typical target is determined.

[0044] In this embodiment of the invention, the similarity is determined as follows: the Euclidean distance between the typical histogram distribution features and the original histogram distribution features is calculated; and the similarity between each target in the overexposed image and its corresponding typical target is evaluated based on the calculation results.

[0045] Specifically, the FPGA receives the typical histogram distribution feature number, indexes the typical histogram distribution feature stored in the corresponding storage module according to the number, calculates the Euclidean distance based on the typical histogram distribution feature and the histogram feature information of each target, and quantitatively evaluates the similarity between each target and its corresponding typical target.

[0046] In step 106, the pixel values ​​of the target in the overexposed image are processed according to the similarity to obtain a clear image for fusion and camouflage with other spectral band images.

[0047] In this embodiment of the invention, a clear image is obtained through the following process: based on the evaluation results, targets with a similarity greater than 95% are identified as successfully matched targets; the typical pixel values ​​corresponding to the successfully matched targets in the typical histogram distribution features are indexed according to the number of the typical histogram distribution features; the original pixels of the successfully matched targets in the overexposed image are replaced with the typical pixel values ​​to obtain a clear image for fusion and camouflage with other spectral band images.

[0048] Specifically, targets with a similarity of over 95% can be considered to match the identified target. The FPGA indexes the pixel values ​​of the corresponding typical targets according to the typical histogram distribution characteristics, and replaces the target pixel values ​​in the overexposed image to obtain a clear image for camouflage evaluation by fusion with other spectral band images.

[0049] It is worth noting that this embodiment uses an FPGA for morphological filtering, target region cropping, and histogram statistics, based on the inherent characteristics of the FPGA. Due to its parallel computing capabilities and ability to build underlying circuits, the FPGA is suitable for pipelined computing and image preprocessing algorithms such as morphological filtering and histogram statistics. The use of an RK3588 for target recognition, target edge extraction, and feature matching is based on the inherent characteristics of the RK3588. The RK3588 has a certain level of computing power that can be used to accelerate AI inference, making it suitable for the fast response requirements of target recognition on embedded platforms. Its computing power can also run algorithms to achieve target edge extraction and feature matching.

[0050] Please refer to Figure 2 This invention provides an overexposed image processing apparatus for fusion camouflage assessment, the apparatus comprising:

[0051] The first processing module 200 is used to perform morphological filtering and AI recognition processing on the acquired overexposed image to obtain the filtered image and the recognition result in turn.

[0052] The first determining module 202 is used to extract the contour of the filtered image and determine the typical histogram distribution features and the original histogram distribution features of the target in sequence according to the extraction result and the recognition result;

[0053] The second determining module 204 is used to determine the similarity between the target in the overexposed image and the corresponding typical target based on the typical histogram distribution characteristics and the original histogram distribution characteristics.

[0054] The second processing module 206 is used to process the pixel values ​​of the target in the overexposed image according to the similarity, so as to obtain a clear image for fusion and camouflage with other spectral band images.

[0055] In this embodiment of the invention, the recognition result includes target location information, target location coordinates, and target classification result in the overexposed image.

[0056] In this embodiment of the invention, the contour extraction of the filtered image includes: determining the minimum connected component coordinates of each target based on the edge contour information of all targets in the filtered image; and processing the number of image pixels occupied by each target based on spatial resolution information to obtain the two-dimensional projected area of ​​each target.

[0057] In this embodiment of the invention, determining the typical histogram distribution features and the original histogram distribution features of the target sequentially based on the extraction results and the recognition results includes: matching the two-dimensional projected area and the recognition results in a histogram distribution feature database to obtain the typical histogram distribution features of the target's most likely corresponding category, and the number of the typical histogram distribution features; extracting target image data based on the coordinates of the minimum connected component, and performing histogram statistics on the target image data to obtain the original histogram distribution features of each target.

[0058] In this embodiment of the invention, determining the similarity between a target in an overexposed image and its corresponding typical target based on the typical histogram distribution features and the original histogram distribution features includes: calculating the Euclidean distance between the typical histogram distribution features and the original histogram distribution features; and evaluating the similarity between each target in the overexposed image and its corresponding typical target based on the calculation results.

[0059] In this embodiment of the invention, the pixel values ​​of the target in the overexposed image are processed according to the similarity to obtain a clear image for fusion camouflage with other spectral band images. This includes: determining targets with a similarity greater than 95% as successfully matched targets based on evaluation results; indexing the typical pixel values ​​corresponding to the successfully matched targets in the typical histogram distribution features according to the number of the typical histogram distribution features; and replacing the original pixels of the successfully matched targets in the overexposed image with the typical pixel values ​​to obtain a clear image for fusion camouflage with other spectral band images.

[0060] It should be noted that the overexposed image processing apparatus for camouflage assessment provided in the above embodiments is only an example of the division of the above functional modules. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the apparatus can be divided into different functional modules to complete all or part of the functions described above. In addition, the overexposed image processing apparatus for camouflage assessment provided in the above embodiments and the overexposed image processing method for camouflage assessment belong to the same concept, and the specific implementation process is detailed in the method embodiments, which will not be repeated here.

[0061] Embodiments of this application also provide a computer device, please refer to... Figure 3 The computer device includes a processor and a memory, the memory storing at least one instruction, at least one program, code set, or instruction set, wherein the at least one instruction, at least one program, code set, or instruction set is loaded and executed by the processor to implement the overexposure image processing method for fusion camouflage assessment provided in the above-described method embodiments.

[0062] Embodiments of this application also provide a computer-readable storage medium storing at least one instruction, at least one program, code set, or instruction set, wherein the at least one instruction, at least one program, code set, or instruction set is loaded and executed by a processor to implement the overexposure image processing method for fusion camouflage evaluation provided in the above-described method embodiments.

[0063] Embodiments of this application also provide a computer program product comprising a computer program, wherein a processor of a computer device reads the computer program from a computer-readable storage medium, and the processor executes the computer program, causing the computer device to perform any of the overexposed image processing methods for fusion camouflage assessment described in the above embodiments.

[0064] For ease of description, the above systems or devices are described separately as various modules or units based on their functions. Of course, in implementing this application, the functions of each unit can be implemented in one or more software and / or hardware components.

[0065] As can be seen from the above description of the embodiments, those skilled in the art can clearly understand that this application can be implemented by means of software plus necessary general-purpose hardware platforms. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in various embodiments or some parts of the embodiments of this application.

[0066] Finally, it should be noted that in this document, relational terms such as first, second, third, and fourth are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0067] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for processing overexposed images for camouflage assessment, characterized in that, The method includes: The obtained overexposed images are subjected to morphological filtering and AI recognition processing respectively to obtain filtered images and recognition results. Contour extraction is performed on the filtered image, and the typical histogram distribution features and original histogram distribution features of the target are determined sequentially based on the extraction and recognition results, including: Based on the edge contour information of all targets in the filtered image, determine the minimum connected component coordinates of each target. The number of image pixels occupied by each target is processed according to the spatial resolution information to obtain the two-dimensional projected area of ​​each target; Based on the two-dimensional projected area and the recognition result, a match is made in the histogram distribution feature library to obtain the typical histogram distribution feature of the target's most likely corresponding category, as well as the number of the typical histogram distribution feature. The target image data is extracted based on the coordinates of the minimum connected component, and histogram statistics are performed on the target image data to obtain the original histogram distribution characteristics of each target. Based on the typical histogram distribution characteristics and the original histogram distribution characteristics, the similarity between the target in the overexposed image and the corresponding typical target is determined; The pixel values ​​of the target in the overexposed image are processed according to the similarity to obtain a clear image for camouflage evaluation by fusion with other spectral band images.

2. The method as described in claim 1, characterized in that, The recognition results include target location information, target location coordinates, and target classification results in the overexposed image.

3. The method as described in claim 1, characterized in that, The step of determining the similarity between a target in an overexposed image and its corresponding typical target based on the typical histogram distribution characteristics and the original histogram distribution characteristics includes: Calculate the Euclidean distance between the typical histogram distribution characteristics and the original histogram distribution characteristics; The similarity between each target in the overexposed image and its corresponding typical target is evaluated based on the calculation results.

4. The method as described in claim 3, characterized in that, The pixel values ​​of the target in the overexposed image are processed according to the similarity to obtain a clear image for camouflage evaluation by fusion with other spectral band images, including: Based on the evaluation results, targets with a similarity greater than 95% are considered successful matches. The typical pixel value corresponding to the successfully matched target in the typical histogram distribution feature is indexed according to the number of the typical histogram distribution feature; The original pixels of the successfully matched target in the overexposed image are replaced with the typical pixel values ​​to obtain a clear image for camouflage evaluation by fusion with other spectral band images.

5. An overexposed image processing apparatus for fusion camouflage assessment, characterized in that, The apparatus, used in the method of any one of claims 1-4, comprises: The first processing module is used to perform morphological filtering and AI recognition processing on the acquired overexposed image to obtain the filtered image and the recognition result in turn. The first determining module is used to extract the contour of the filtered image and determine the typical histogram distribution features and the original histogram distribution features of the target in sequence based on the extraction results and the recognition results. The second determining module is used to determine the similarity between the target in the overexposed image and the corresponding typical target based on the typical histogram distribution characteristics and the original histogram distribution characteristics. The second processing module is used to process the pixel values ​​of the target in the overexposed image according to the similarity level to obtain a clear image for camouflage evaluation by fusion with other spectral band images.

6. A computer device, characterized in that, The computer device includes a memory and a processor. The memory is used to store computer programs, and the processor is used to execute the computer programs stored in the memory to implement the steps of the method according to any one of claims 1-4.

7. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, which, when executed by a processor, implements the steps of the method described in any one of claims 1-4.

8. A computer program product, characterized in that, Includes a computer program, which, when executed by a processor, implements the steps of the method according to any one of claims 1-4.

Images