Target enhancement method and device based on high-time-phase image

Abstract

The invention discloses a target enhancement method and device based on a high-time-phase image. The method comprises the following steps: filtering a plurality of input original sequential images to obtain filtered sequential images; the original sequential image is taken from a high-time image acquired by high-time-phase equipment; standardizing the filtered sequential image to obtain a standardized sequential image; performing sliding mapping on the standardized sequential image according to a preset multi-core combination kernel function to obtain a sequential mapping image; standardizing the sequential mapping image to obtain a standardized sequential mapping image; and taking a maximum value in the standardized sequential mapping image pixel by pixel to obtain an enhanced image. According to the invention, the target in the image can be effectively enhanced, and the identification capability of the image is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and in particular to a method and device for target enhancement based on high-phase images. Background Art

[0002] In computer image analysis, the purpose of target enhancement is to make objects of interest, such as vehicles, pedestrians, obstacles, etc., clearer and more prominent in the image through image processing technology. Since images obtained in practical applications such as remote sensing, security monitoring, and unmanned driving are usually affected by factors such as low resolution, strong noise, and complex background environment, the difference between the target and the background becomes unclear, increasing the difficulty of subsequent target detection and recognition. Therefore, various target enhancement algorithms are needed to make objects of interest clearer and more prominent, so that these targets can be detected and recognized more easily.

[0003] Related technologies are usually based on single-phase image target enhancement algorithms, which mainly rely on the spatial domain information of the image for target enhancement. However, since such algorithms are too dependent on image information at a single moment, the target enhancement effect is poor when faced with low image resolution, strong noise, complex background environment, etc., which can easily lead to misdetection or loss of targets.

[0004] Therefore, there is an urgent need for a target enhancement method and device based on high-phase images. Summary of the invention

[0005] The present invention provides a method and device for target enhancement based on high-phase images, which can effectively enhance the target in the image and improve the image recognition ability. The technical solution is as follows:

[0006] On the one hand, a method for object enhancement based on a high-phase image is provided, the method comprising:

[0007] Performing filtering processing on multiple input original sequential images to obtain filtered sequential images; the original sequential images are taken from high-time images obtained by a high-time phase device;

[0008] Normalizing the filtered sequential image to obtain a standardized sequential image;

[0009] Perform sliding mapping on the standardized sequential image according to a preset multi-core combined kernel function to obtain a sequential mapping image;

[0010] Standardizing the sequential mapping image to obtain a standardized sequential mapping image;

[0011] The maximum value of the standardized sequential mapping image is taken pixel by pixel to obtain the enhanced image.

[0012] On the other hand, a device for target enhancement based on high-phase images is provided, the device comprising:

[0013] A filtering unit, used for filtering the inputted multiple original sequential images to obtain filtered sequential images; the original sequential images are taken from high-time images acquired by a high-time phase device;

[0014] A first standardization unit is used to standardize the filtered sequential image to obtain a standardized sequential image;

[0015] A mapping unit, used for performing sliding mapping on the standardized sequential image according to a preset multi-core combined kernel function to obtain a sequential mapping image;

[0016] A second standardization unit is used to standardize the sequential mapping image to obtain a standardized sequential mapping image;

[0017] The determination unit is used to take the maximum value of the standardized sequential mapping image pixel by pixel to obtain the enhanced image.

[0018] On the other hand, a computer device is provided, which includes a memory and a processor, wherein the memory is used to store a computer program, and the processor is used to execute the computer program stored in the memory to implement the steps of the above-mentioned target enhancement method based on high-phase images.

[0019] On the other hand, a computer-readable storage medium is provided, wherein a computer program is stored in the storage medium, and when the computer program is executed by a processor, the steps of the above-mentioned target enhancement method based on high-phase images are implemented.

[0020] On the other hand, a computer program product is provided, comprising a computer program, wherein when the computer program is executed by a processor, the steps of the above-mentioned object enhancement method based on high temporal phase images are implemented.

[0021] The embodiment of the present invention provides a method for target enhancement based on high-phase images. When performing target enhancement, a sequential image taken from a high-phase image is used as input, and the input sequential image is filtered, standardized, and slidingly mapped in turn to obtain a sequential mapping image. The sequential mapping image is then standardized, and the maximum value after standardization is used as the enhanced image. In this way, the regularity information of the spatial features in the high-phase image changing over time can be utilized, and by accumulating and analyzing information in the time dimension, it is possible to effectively enhance the target in the image and improve the image recognition capability under conditions of low image resolution, strong noise, and complex background environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0023] Figure 1 is a flow chart of a method for target enhancement based on high-phase images provided by one embodiment of the present invention;

[0024] Figure 2 is a structural diagram of a target enhancement device based on high-phase images provided by an embodiment of the present invention;

[0025] Figure 3 is a hardware architecture diagram of a computer device provided by an embodiment of the present invention;

[0026] Figures 4 to 14 They are respectively images at different times acquired by using a high temporal phase device provided by an embodiment of the present invention;

[0027] Fig.15 This is an enhanced image obtained by using the method of the present application and provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0028] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0029] As mentioned above, the related technology mainly performs target enhancement based on single-phase images. The inventors found in their work that the algorithm has the following main defects: First, when facing targets with occlusion or fine structures, the limited spatial domain information makes the enhancement effect unsatisfactory, which easily leads to the target being ignored or misdetected. Secondly, noise is the main problem in single-phase processing. Although filtering and other technologies can reduce noise, in the case of severe noise, the algorithm may not be able to effectively maintain the details of the target, resulting in blurred or lost target information. Finally, when the background environment is more complex and the contrast is low, it is more difficult to distinguish between the target and the background, and the target enhancement algorithm based on single-phase images may also fail.

[0030] Based on the above problems, the inventors proposed that the regular information of spatial features in high-phase images changing over time can be utilized to accumulate and analyze information in the time dimension, thereby effectively enhancing the targets in the image and improving the image recognition capability in cases of low image resolution, strong noise, and complex background environment.

[0031] The specific implementation of the above concept is described below.

[0032] Please refer to Figure 1 , an embodiment of the present invention provides a method for target enhancement based on high-phase images, the method comprising:

[0033] Step 100, filtering the input multiple original sequential images to obtain filtered sequential images; the original sequential images are taken from high-time images obtained by a high-time phase device;

[0034] Step 102, normalizing the filtered sequential image to obtain a standardized sequential image;

[0035] Step 104, performing sliding mapping on the standardized sequential image according to a preset multi-core combined kernel function to obtain a sequential mapping image;

[0036] Step 106, normalizing the sequential mapping image to obtain a standardized sequential mapping image;

[0037] Step 108 , taking the maximum value of the standardized sequential mapping image pixel by pixel to obtain an enhanced image.

[0038] In an embodiment of the present invention, when performing target enhancement, a sequential image taken from a high-time image is used as input, and the input sequential image is filtered, normalized, and slidingly mapped in turn to obtain a sequential mapping image. The sequential mapping image is then standardized, and the maximum value after normalization is used as the enhanced image. In this way, the regularity information of the spatial features in the high-time image changing over time can be utilized, and by accumulating and analyzing information in the time dimension, the target in the image can be effectively enhanced and the image recognition capability can be improved under the conditions of low image resolution, strong noise, complex background environment, etc.

[0039] Described below Figure 1 How the various steps are performed.

[0040] First, with respect to step 100, a plurality of original sequential images are filtered to obtain filtered sequential images, including:

[0041] The first formula is used to perform time domain sliding average filtering on the input multiple original sequential images to obtain filtered sequential images;

[0042] The expression of the first formula is:

[0043]

[0044] In the formula, represents the image after time domain sliding average filtering; p represents the original image; 2n+1 is the sliding average window length; n is a positive integer, -n and n represent the starting position and the end position of the sliding window respectively; i, j represent the distance and azimuth of the image respectively; k represents the time domain position of the image; l represents the time domain position of the sliding window.

[0045] In this embodiment, k and l are variables in the time domain, and k can be understood as the number of frames of images. For example, when k=10 and n=5, k+l represents the 5th to 15th frames of images. This step can effectively filter out noise interference by filtering the original image.

[0046] With respect to step 102, the filtered sequential image is subjected to block Z-score normalization using the second formula to obtain a normalized sequential image;

[0047] The expression of the second formula is:

[0048]

[0049] In the formula, represents the normalized sequential images; represents the image after sliding average filtering in the time domain, m seg and σ seg They represent the mean and standard deviation of all pixels in the sequential block image respectively; seg{·} represents the block operation; sti{·} represents the splicing operation; i, j represent the distance and azimuth of the image respectively; k represents the time domain position of the image.

[0050] In this embodiment, block Z-score normalization can prevent faint targets around strong targets from being masked, and maximize the retention of detail information in sequential images.

[0051] For step 104, the mapping relationship of the multi-core combined kernel function is:

[0052]

[0053] In the formula, k b (x,y) represents the multi-core mapping function; x m is the element in the previous signal segment; y m is the element in the subsequent signal segment; η is the modulation coefficient; B is the total length of the signal; L is the sliding window length; I is the sliding window interval; b represents the position of the high-dimensional mapping result along the time series; B-2L-I+1 is the total length after mapping.

[0054] In this embodiment, according to the mapping relationship, for each pixel amplitude time series, two sliding windows with a length of L and an interval of I are set from the starting position, and the amplitude segments in the window are sorted by intensity. The window is slid with a preset step size in the time sequence to calculate the similarity of the two amplitude segments in the window, and the cycle is repeated until all images are traversed to obtain a sequential mapping image.

[0055] For step 106, the sequential mapping image is normalized using the second formula to obtain a normalized sequential mapping image. Similarly, the block Z-score normalization can avoid the masking of weak targets around strong targets and maximize the retention of detail information in the sequential mapping image.

[0056] Finally, for step 108, the enhanced image is determined by the following formula, that is, the process of taking the maximum value can be described as:

[0057] f(i,j)=max{r(i,j,b)}b=1,2,…,(B-2L-I+1)

[0058] Where f is the enhanced image; r is the normalized sequential mapping image; max{·} is the maximum value operation; i, j represent the distance and azimuth of the image, respectively; B is the total signal length; L is the sliding window length; I is the sliding window interval; b represents the time sequence position of the high-dimensional mapping result; B-2L-I+1 is the total length after mapping.

[0059] The enhanced image determined in this step can be used for subsequent target detection and intelligent recognition, target tracking and trajectory extraction.

[0060] In order to prove the effectiveness of the method of this application, the inventor Figures 4 to 14 The high-phase sequential image shown in the figure is used as input, and the target enhancement processing is performed using the method of the present application to obtain Fig.15 The target image is shown. Fig.15 It can be seen from the processing results that the noise and background environment in the input sequential images are suppressed, and the continuously moving targets are enhanced. The above results illustrate the effectiveness of the target enhancement method based on high-phase images of the present application.

[0061] like Figure 2 , Figure 3 As shown, an embodiment of the present invention provides a target enhancement device based on high-phase images. The device embodiment can be implemented by software, or by hardware or a combination of software and hardware. From the hardware level, Figure 2 As shown in FIG. 1 , a hardware architecture diagram of a computing device where a target enhancement device based on high-phase images is provided in an embodiment of the present invention is located. Figure 2In addition to the processor, memory, network interface, and non-volatile memory shown in the figure, the computing device in which the device is located in the embodiment may also generally include other hardware, such as a forwarding chip responsible for processing messages, etc. Taking software implementation as an example, Figure 3 As shown, as a device in a logical sense, the CPU of the computing device in which it is located reads the corresponding computer program in the non-volatile memory into the internal memory and runs it.

[0062] Please refer to Figure 3 The embodiment of the present application provides a target enhancement device based on a high-phase image, comprising:

[0063] The filtering unit 300 is used to filter the input multiple original sequential images to obtain filtered sequential images; the original sequential images are taken from the high-time images obtained by the high-time phase device;

[0064] A first standardization unit 302 is used to standardize the filtered sequential image to obtain a standardized sequential image;

[0065] A mapping unit 304 is used to perform sliding mapping on the standardized sequential image according to a preset multi-core combined kernel function to obtain a sequential mapping image;

[0066] A second standardization unit 306 is used to standardize the sequential mapping image to obtain a standardized sequential mapping image;

[0067] The determination unit 308 is used to obtain the maximum value of the standardized sequential mapping image pixel by pixel to obtain an enhanced image.

[0068] In some implementations, filtering is performed on the input multiple original sequential images to obtain filtered sequential images, including:

[0069] The first formula is used to perform time domain sliding average filtering on the input multiple original sequential images to obtain filtered sequential images;

[0070] The expression of the first formula is:

[0071]

[0072] In the formula, represents the image after time domain sliding average filtering; p represents the original image; 2n+1 is the sliding average window length; n is a positive integer, -n and n represent the starting position and the end position of the sliding window respectively; i, j represent the distance and azimuth of the image respectively; k represents the time domain position of the image; l represents the time domain position of the sliding window.

[0073] In some embodiments, the filtered sequential image is subjected to block Z-score normalization using the second formula to obtain a normalized sequential image;

[0074] The expression of the second formula is:

[0075]

[0076] In the formula, represents the normalized sequential images; represents the image after sliding average filtering in the time domain, m seg and σ seg They represent the mean and standard deviation of all pixels in the sequential block image respectively; seg{·} represents the block operation; sti{·} represents the splicing operation; i, j represent the distance and azimuth of the image respectively; k represents the time domain position of the image.

[0077] In some implementations, the mapping relationship of the multi-core combined kernel function is:

[0078]

[0079] In the formula, k b (x,y) represents the multi-core mapping function; x m is the element in the previous signal segment; y m is the element in the subsequent signal segment; η is the modulation coefficient; B is the total length of the signal; L is the sliding window length; I is the sliding window interval; b represents the position of the high-dimensional mapping result along the time series; B-2L-I+1 is the total length after mapping.

[0080] In some implementations, the second formula is used to normalize the sequential mapping image to obtain a normalized sequential mapping image.

[0081] In some embodiments, the enhanced image is determined by the following formula:

[0082] f(i,j)=max{r(i,j,b)}b=1,2,…,(B-2L-I+1)

[0083] Where f is the enhanced image; r is the normalized sequential mapping image; max{·} is the maximum value operation; i, j represent the distance and azimuth of the image, respectively; B is the total signal length; L is the sliding window length; I is the sliding window interval; b represents the time sequence position of the high-dimensional mapping result; B-2L-I+1 is the total length after mapping.

[0084] It should be noted that the target enhancement device based on high-phase images provided in the above embodiment is only illustrated by the division of the above-mentioned functional modules. In actual applications, the above-mentioned functions can be assigned to different functional modules as needed, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In addition, the target enhancement device based on high-phase images provided in the above embodiment and the target enhancement method embodiment based on high-phase images belong to the same concept. The specific implementation process is detailed in the method embodiment and will not be repeated here.

[0085] The embodiment of the present application also provides a computer device, please refer to Figure 3 The computer device includes a processor and a memory, in which at least one instruction, at least one program, a code set or an instruction set is stored, and the at least one instruction, at least one program, a code set or an instruction set is loaded and executed by the processor to implement the target enhancement method based on high-phase images provided in the above-mentioned method embodiments.

[0086] An embodiment of the present application also provides a computer-readable storage medium, on which is stored at least one instruction, at least one program, code set or instruction set, and the at least one instruction, at least one program, code set or instruction set is loaded and executed by a processor to implement the target enhancement method based on high phase images provided in the above-mentioned method embodiments.

[0087] An embodiment of the present application also provides a computer program product, which includes a computer program. A processor of a computer device reads the computer program from a computer-readable storage medium, and the processor executes the computer program, so that the computer device executes any target enhancement method based on high-phase images described in the above embodiments.

[0088] For the convenience of description, the above system or device is described by dividing it into various modules or units according to its functions. Of course, when implementing the present application, the functions of each unit can be implemented in the same or multiple software and / or hardware.

[0089] It can be known from the description of the above implementation methods that those skilled in the art can clearly understand that the present application can be implemented by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the present application can be essentially or partly contributed to the prior art in the form of a software product, which can be stored in a storage medium such as ROM / RAM, a magnetic disk, an optical disk, etc., and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in the various embodiments of the present application or certain parts of the embodiments.

[0090] Finally, it should be noted that, in this article, relational terms such as first, second, third and fourth 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 "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0091] The above is only a preferred implementation of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A target enhancement method based on high-phase image, characterized in that: The method comprises: Performing filtering processing on multiple input original sequential images to obtain filtered sequential images; the original sequential images are taken from high-time images obtained by a high-time phase device; Normalizing the filtered sequential image to obtain a standardized sequential image; Perform sliding mapping on the standardized sequential image according to a preset multi-core combined kernel function to obtain a sequential mapping image; Standardizing the sequential mapping image to obtain a standardized sequential mapping image; The maximum value of the standardized sequential mapping image is taken pixel by pixel to obtain the enhanced image.

2. The method according to claim 1, characterized in that The filtering process is performed on the input multiple original sequential images to obtain filtered sequential images, including: The first formula is used to perform time domain sliding average filtering on the input multiple original sequential images to obtain filtered sequential images; The expression of the first formula is: In the formula, represents the image after time domain sliding average filtering; p represents the original image; 2n+1 is the sliding average window length; n is a positive integer, -n and n represent the starting position and the end position of the sliding window respectively; i, j represent the distance and azimuth of the image respectively; k represents the time domain position of the image; l represents the time domain position of the sliding window.

3. The method according to claim 1, characterized in that The second formula is used to perform block Z-score normalization on the filtered sequential image to obtain a normalized sequential image; The expression of the second formula is: In the formula, represents the normalized sequential images; represents the image after sliding average filtering in the time domain, m seg and σ seg They represent the mean and standard deviation of all pixels in the sequential block image respectively; seg{·} represents the block operation; sti{·} represents the splicing operation; i, j represent the distance and azimuth of the image respectively; k represents the time domain position of the image.

4. The method according to claim 1, characterized in that: The mapping relationship of the multi-core combined kernel function is: In the formula, k b (x,y) represents the multi-core mapping function; x m is the element in the previous signal segment; y m is the element in the post-signal segment; η is the modulation coefficient; B is the total signal length; L is the sliding window length; I is the sliding window interval; b represents the time sequence position of the high-dimensional mapping result; B-2L-I+1 is the total length after mapping.

5. The method according to claim 3, characterized in that: The second formula is used to standardize the sequential mapping image to obtain a standardized sequential mapping image.

6. The method according to claim 4, characterized in that The enhanced image is determined by the following formula: f(i,j)=max{r(i,j,b)}b=1,2,…,(B-2L-I+1) Where f is the enhanced image; r is the normalized sequential mapping image; max{·} is the maximum value operation; i, j represent the distance and azimuth of the image, respectively; B is the total signal length; L is the sliding window length; I is the sliding window interval; b represents the time sequence position of the high-dimensional mapping result; B-2L-I+1 is the total length after mapping.

7. A target enhancement device based on high-phase images, characterized in that: The device comprises: A filtering unit, used for filtering the inputted multiple original sequential images to obtain filtered sequential images; the original sequential images are taken from high-time images acquired by a high-time phase device; A first standardization unit is used to standardize the filtered sequential image to obtain a standardized sequential image; A mapping unit, used for performing sliding mapping on the standardized sequential image according to a preset multi-core combined kernel function to obtain a sequential mapping image; A second standardization unit is used to standardize the sequential mapping image to obtain a standardized sequential mapping image; The determination unit is used to take the maximum value of the standardized sequential mapping image pixel by pixel to obtain the enhanced image.

8. A computer device, characterized in that: The computer device includes a memory and a processor, the memory is used to store a computer program, and the processor is used to execute the computer program stored in the memory to implement the steps of any one of the methods described in claims 1-6.

9. A computer-readable storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method described in any one of claims 1 to 6 are implemented.

10. A computer program product, characterized in that The method comprises a computer program, wherein when the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images