Bottom echo image processing method and device, electronic equipment and storage medium

Abstract

The application provides a bottom echo image processing method and device, electronic equipment and storage medium, by acquiring an original bottom echo signal, performing signal domain processing on the original bottom echo signal, and constructing a preliminary bottom echo image; performing image domain processing on the preliminary bottom echo image, including: extracting an image structure information graph of the preliminary bottom echo image, searching for similar image blocks in the preliminary bottom echo image based on the image structure information graph, and combining the searched similar image blocks into a three-dimensional block group; transforming the three-dimensional group to a transform domain for coefficient contraction processing, and generating a bottom echo image through inverse transformation and weighted aggregation, the application can effectively suppress various noise interferences introduced by a complex marine environment, significantly improve the signal-to-noise ratio and data quality of the bottom echo image, can accurately retain the edge details and weak target information of the image while correcting the intensity anomaly of the central beam area, and provides reliable data support for high-quality seabed sonar image acquisition, bottom quality classification and target detection.

Description

Technical Field

[0001] This invention relates to the field of geodesy and surveying engineering technology, and in particular to a bottom echo image processing method, apparatus, electronic device and storage medium. Background Technology

[0002] The ocean is rich in biological, mineral, and energy resources, making it a key area for strategic development for countries worldwide. As a fundamental support for marine economic development, the importance of marine exploration technology continues to rise. Multibeam bathymetry (MBB) systems, with their advantages of high precision, full coverage, and high efficiency, have become the mainstream technology for marine mapping and seabed topography surveys, capable of simultaneously acquiring multiple types of data, including water depth, water volume, and bottom echo intensity. Among these, the seabed echo image generated after imaging processing of bottom echo intensity data is a crucial basis for seabed sediment classification, target detection, resource exploration, and marine engineering site selection. However, due to the dual constraints of the complex marine environment and the imaging mechanism of multibeam sonar, current seabed echo imaging still faces two major technical challenges. One is the problem of intensity anomalies in the central beam region. Multibeam sonar exhibits significant differences in reflection patterns at different incident angles, with the central beam region exhibiting predominantly specular reflection, resulting in significantly higher echo intensity than the peripheral beam regions. While existing preprocessing methods correct for factors such as propagation loss, acoustic illumination area, and topographic relief, they still struggle to completely eliminate intensity imbalances between the central and non-central beam regions. This results in striped intensity anomalies in bottom echo images, severely impacting image uniformity and subsequent interpretation accuracy. Secondly, there is the problem of complex noise interference. During operations, bottom echo intensity data inevitably contains marine environmental noise, platform self-noise, and seabed reverberation noise. These noises overlap with the effective signal in both the time and frequency domains. Traditional filtering methods such as median filtering and Wiener filtering struggle to effectively preserve edge details and weak target information in the echo signal while suppressing noise, leading to low signal-to-noise ratios and poor clarity in bottom echo images, failing to meet the application requirements for high-precision seabed classification and small target detection. Summary of the Invention

[0003] This invention provides a bottom echo image processing method, apparatus, electronic device, and storage medium to solve the problem that existing technologies cannot simultaneously solve the problem of abnormal intensity in the central beam region and complex marine noise interference in multibeam bottom echo images. Furthermore, these technologies cannot effectively preserve edge details and weak target information when suppressing noise, resulting in image quality that is difficult to meet the requirements of high-precision seabed exploration applications.

[0004] This invention provides a method for processing low-echo images, comprising: The raw bottom echo signal is acquired, and the raw bottom echo signal is processed in the signal domain to construct a preliminary bottom echo image. Image domain processing of the preliminary bottom echo image includes: extracting the image structure information map of the preliminary bottom echo image; searching for similar image blocks in the preliminary bottom echo image based on the image structure information map; combining the searched similar image blocks into a three-dimensional block group; transforming the three-dimensional group to the transform domain for coefficient shrinkage processing; and generating a bottom echo image through inverse transformation and weighted aggregation.

[0005] According to the bottom echo image processing method provided by the present invention, the step of performing signal domain processing on the original bottom echo signal to construct a preliminary bottom echo image includes: Empirical mode decomposition is performed on the original bottom echo signal to obtain several intrinsic mode function components and residual components; Calculate the correlation coefficient between each intrinsic mode function component and the original bottom echo signal, and divide each intrinsic mode function component into high-frequency components and low-frequency components based on the correlation coefficient; The high-frequency components are subjected to wavelet threshold denoising, while the low-frequency components and residual components are subjected to noise component zeroing. The high-frequency components that have undergone noise reduction, the low-frequency components that have been zeroed out, and the residual components are reconstructed to obtain the noise-reduced low echo signal. The noise-reduced bottom echo signals corresponding to each beam position are superimposed to generate a preliminary bottom echo image.

[0006] According to the bottom echo image processing method provided by the present invention, the step of extracting the image structure information map of the preliminary bottom echo image includes: Edge detection is performed on the preliminary bottom echo image, and the edge detection results are used as the image structure information map.

[0007] According to the bottom echo image processing method provided by the present invention, the step of searching for similar image blocks in the preliminary bottom echo image based on the image structure information map, and combining the searched similar image blocks into a three-dimensional block group, includes: A reference image block is selected from the preliminary bottom echo image; Based on the image structure information map, the similarity metric between the candidate image block and the reference image block is calculated within the search window, and the candidate image block that meets the preset similarity condition is taken as a similar image block. The reference image block and its corresponding similar image block are stacked along the third dimension to form a three-dimensional block group.

[0008] According to the bottom echo image processing method provided by the present invention, the step of transforming the three-dimensional group to the transform domain for coefficient shrinkage processing includes: A two-dimensional transformation is performed on each image block in the three-dimensional block group, and a one-dimensional transformation is performed on the stacking dimension of the three-dimensional block group to obtain the three-dimensional transformation coefficients. Threshold filtering is applied to the three-dimensional transformation coefficients, retaining those greater than a preset threshold and setting coefficients not greater than the preset threshold to zero; The preset threshold is determined based on the noise level estimate of the high-frequency components in the transform domain.

[0009] According to the bottom echo image processing method provided by the present invention, the inverse transform includes: The three-dimensional transformation coefficients that are greater than the preset threshold are subjected to a one-dimensional inverse transformation along the stacking dimension, and a two-dimensional inverse transformation is performed on each image block to obtain the denoised three-dimensional block group estimate.

[0010] According to the bottom echo image processing method provided by the present invention, the weighted aggregation includes: Determine the weight of each pixel within each image block in each of the three-dimensional block group estimates; For each pixel in the preliminary bottom echo image, the estimated values ​​of all image blocks containing that pixel at that pixel and their corresponding weights are obtained, and a weighted average is performed to obtain the final pixel value of that pixel. A bottom echo image is generated based on the final pixel values ​​of all pixels.

[0011] The present invention also provides a bottom echo image processing apparatus, comprising: The signal domain processing module is used to acquire the original bottom echo signal, perform signal domain processing on the original bottom echo signal, and construct a preliminary bottom echo image. The image domain processing module is used to perform image domain processing on the preliminary bottom echo image, including: extracting the image structure information map of the preliminary bottom echo image; searching for similar image blocks in the preliminary bottom echo image based on the image structure information map; combining the searched similar image blocks into a three-dimensional block group; transforming the three-dimensional group to the transform domain for coefficient shrinkage processing; and generating a bottom echo image through inverse transformation and weighted aggregation.

[0012] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the bottom echo image processing method as described in any of the preceding claims.

[0013] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the bottom echo image processing method described in any of the preceding claims.

[0014] The bottom echo image processing method, apparatus, electronic device, and storage medium provided by this invention acquire raw bottom echo signals, perform signal domain processing on the raw bottom echo signals to construct a preliminary bottom echo image; perform image domain processing on the preliminary bottom echo image, including: extracting an image structure information map of the preliminary bottom echo image; searching for similar image blocks in the preliminary bottom echo image based on the image structure information map; combining the searched similar image blocks into a three-dimensional block group; transforming the three-dimensional group to the transform domain for coefficient shrinkage processing; and generating a bottom echo image through inverse transformation and weighted aggregation. This invention, by adaptively decomposing and differentially denoising the raw echo in the signal domain, can effectively suppress various noise interferences introduced by complex marine environments, significantly improving the signal-to-noise ratio and data quality of the bottom echo image; furthermore, by introducing structure-prior-guided similar block matching and three-dimensional collaborative filtering in the image domain, it can accurately preserve the edge details and weak target information of the image while correcting the intensity anomaly in the central beam region, providing reliable data support for high-precision seabed sediment classification and target detection. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in this 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 this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0016] Figure 1 This is one of the flowcharts of the bottom echo image processing method provided in the embodiments of the present invention; Figure 2 This is an IMF component map provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of edge detection provided in an embodiment of the present invention; Figure 4 This is the second flowchart of the bottom echo image processing method provided in the embodiments of the present invention; Figure 5 This is a functional structure diagram of the bottom echo image processing device provided in an embodiment of the present invention; Figure 6 This is a functional structure diagram of the electronic device provided in the embodiments of the present invention. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0018] Figure 1 The flowchart of the bottom echo image processing method provided in the embodiments of the present invention is as follows: Figure 1 As shown, the bottom echo image processing method provided in this embodiment of the invention includes: Step 101: Obtain the original bottom echo signal, perform signal domain processing on the original bottom echo signal, and construct a preliminary bottom echo image; Step 102: Perform image domain processing on the preliminary bottom echo image, including: extracting the image structure information map of the preliminary bottom echo image; searching for similar image blocks in the preliminary bottom echo image based on the image structure information map; combining the searched similar image blocks into a three-dimensional block group; transforming the three-dimensional group to the transform domain for coefficient shrinkage processing; and generating a bottom echo image through inverse transformation and weighted aggregation.

[0019] Traditional bottom echo image processing faces challenges such as intensity anomalies in the central beam region and complex noise interference. While existing preprocessing methods correct for factors like propagation loss, acoustic illumination area, and terrain undulation, they still struggle to completely eliminate intensity imbalances between the central and non-central beam regions. This results in striped intensity anomalies in bottom echo images, severely impacting image uniformity and subsequent interpretation accuracy. Traditional filtering methods, such as median filtering and Wiener filtering, often fail to effectively preserve edge details and weak target information in the echo signal while suppressing noise, leading to low signal-to-noise ratios and poor clarity in bottom echo images, which cannot meet the application requirements of high-precision seabed classification and small target detection.

[0020] The bottom echo image processing method provided in this invention acquires the original bottom echo signal, performs signal domain processing on the original bottom echo signal to construct a preliminary bottom echo image, and performs image domain processing on the preliminary bottom echo image, including: extracting the image structure information map of the preliminary bottom echo image, searching for similar image blocks in the preliminary bottom echo image based on the image structure information map, and combining the searched similar image blocks into a three-dimensional block group; transforming the three-dimensional group to the transform domain for coefficient shrinkage processing, and generating a bottom echo image through inverse transformation and weighted aggregation. This invention, by adaptively decomposing and differentially denoising the original echo in the signal domain, can effectively suppress various noise interferences introduced by complex marine environments, significantly improving the signal-to-noise ratio and data quality of the bottom echo image; furthermore, by introducing structure-prior-guided similar block matching and three-dimensional collaborative filtering in the image domain, it can accurately preserve the edge details and weak target information of the image while correcting the intensity anomaly in the central beam region, providing reliable data support for high-precision seabed sediment classification and target detection.

[0021] Based on any of the above embodiments, the step of performing signal domain processing on the original bottom echo signal to construct a preliminary bottom echo image includes: Step 201: Perform empirical mode decomposition on the original bottom echo signal to obtain several intrinsic mode function (IMF) components and residual components; In this embodiment of the invention, empirical mode decomposition is performed on each ping of the raw bottom echo data acquired by multibeam sonar to obtain the high-frequency and low-frequency intrinsic mode functions of the echo signal sequence.

[0022] like Figure 2 As shown, the empirical mode decomposition method is as follows: In the formula, n is the number of decomposition layers. imf and r n These are the component and residual classifications obtained from IMF decomposition. θ Let θ be the incident angle of the beam, and BS(θ) be the echo intensity value corresponding to the beam with an incident angle of θ.

[0023] Step 202: Calculate the correlation coefficient between each intrinsic mode function component and the original bottom echo signal, and divide each intrinsic mode function component into high-frequency components and low-frequency components according to the correlation coefficient; In this embodiment of the invention, different processing is applied to high-frequency and low-frequency IMFs. The method for determining high-frequency and low-frequency IMFs is as follows: in, RThe correlation coefficient is BS(θ); BS(θ) is the echo intensity corresponding to the beam with an incident angle of θ. This represents the average echo intensity for different incident angles θ (θ∈[0,M]); IMF i The IMF component corresponding to BS(θ) For different incident angles θ IMF i The average value.

[0024] like A value close to 1 indicates that the component is highly correlated with the original signal and is a major signal component; if... A value close to 0 indicates that the component has little correlation with the original signal and contains noise. A threshold can be set to address this. (The value is generally taken as 0.3~0.5). Components less than the threshold are high-frequency components containing noise, while components greater than the threshold are low-frequency components containing the main information.

[0025] Step 203: Perform wavelet threshold denoising on the high-frequency components, and simultaneously set the noise components of the low-frequency components and residual components to zero. The high-frequency IMF needs to be thresholded for noise reduction, and then reconstructed with the low-frequency IMF signal after the residual components are set to zero to obtain the noise-reduced single-ping echo sequence signal.

[0026] In this embodiment of the invention, wavelet transform is performed on the high-frequency IMF components to obtain each wavelet coefficient and its corresponding sub-band. The high-frequency sub-band is then subjected to improved threshold denoising and reconstructed by inverse wavelet transform with the low-frequency sub-band to obtain the high-frequency IMF components.

[0027] The decomposition result obtained by discrete wavelet transform of the high-frequency IMF components consists of several detail coefficients and one approximation coefficient: in, imf ( θ () represents the decomposed value; These are detail coefficients at different scales, i.e., the high-frequency components of the signal, which include noise and some signal details; These are the approximation coefficients for the last layer, which are the low-frequency components of the signal and contain the main information of the signal.

[0028] Step 204: Reconstruct the high-frequency components after noise reduction, the low-frequency components after noise zeroing, and the residual components to obtain the noise-reduced low-echo signal. In this embodiment of the invention, the denoised multibeam bottom echo signal sequence is obtained by reconstructing the denoised high-frequency IMF component and the low-frequency IMF component.

[0029] Step 205: Superimpose the noise-reduced bottom echo signals corresponding to each beam position to generate a preliminary bottom echo image.

[0030] In this embodiment of the invention, the noise-reduced ping echo signals are superimposed to generate a preliminarily noise-reduced multibeam echo image.

[0031] Based on any of the above embodiments, the step of extracting the image structure information map of the preliminary bottom echo image includes: Edge detection is performed on the preliminary bottom echo image, and the edge detection results are used as the image structure information map.

[0032] In this embodiment of the invention, edge detection is performed on the input noisy echo image to obtain an edge information map, such as... Figure 3 As shown.

[0033] Based on any of the above embodiments, the step of searching for similar image patches in the preliminary bottom echo image based on the image structure information map, and combining the searched similar image patches into a three-dimensional block group, includes: Step 301: Select a reference image block from the preliminary bottom echo image; Step 302: Calculate the similarity metric between the candidate image block and the reference image block within the search window based on the image structure information map, and take the candidate image block that meets the preset similarity condition as a similar image block; Search the search window for image patches similar to the current image patch and group them into a 3D image patch group. The similarity measure is usually Euclidean distance. Where d(Br, Bi) is the obtained reference image patch. With candidate tiles The Euclidean distance between them; and They are respectively and The echo intensity value or grayscale value of the corresponding pixel. Step 303: Stack the reference image block and its corresponding similar image block along the third dimension to form a three-dimensional block group.

[0034] Based on any of the above embodiments, the step of transforming the three-dimensional group to the transform domain and performing coefficient shrinkage includes: Step 401: Perform a two-dimensional transformation on each image block in the three-dimensional block group, and perform a one-dimensional transformation on the stacking dimension of the three-dimensional block group to obtain the three-dimensional transformation coefficients; In this embodiment of the invention, the three-dimensional image block group is mapped to the transform domain. The three-dimensional cooperative transformation is divided into a two-dimensional discrete cosine transform (DCT) for each block and a one-dimensional Hadamard transform (HT) for the stacking dimension of similar blocks. Where B is the image patch, This represents the two-dimensional discrete cosine transform result of image patch B; G is the stacking dimension of similar patches. This represents the one-dimensional Hadamard transformation result of G.

[0035] Step 402: Perform threshold filtering on the three-dimensional transformation coefficients, retain the three-dimensional transformation coefficients that are greater than a preset threshold, and set the coefficients that are not greater than the preset threshold to zero; The preset threshold is determined based on the noise level estimate of the high-frequency components in the transform domain.

[0036] In this embodiment of the invention, a threshold is set to separate noise and information components in the transform domain. The noise error is generally estimated using the absolute median difference of the highest frequency component after the transform. in, This is the noise error. t i For the components in the transform domain, media () is the median-finding function.

[0037] Based on any of the above embodiments, the inverse transformation includes: The three-dimensional transformation coefficients that are greater than the preset threshold are subjected to a one-dimensional inverse transformation along the stacking dimension, and a two-dimensional inverse transformation is performed on each image block to obtain the denoised three-dimensional block group estimate.

[0038] Based on any of the above embodiments, the weighted aggregation includes: Step 501: Determine the weight of each pixel in each image block in each of the three-dimensional block group estimates; Step 502: For each pixel in the preliminary bottom echo image, obtain the estimated values ​​of all image blocks containing that pixel at that pixel and their corresponding weights, and perform a weighted average to obtain the final pixel value of that pixel. Step 503: Generate a bottom echo image based on the final pixel values ​​of all pixels.

[0039] In this embodiment of the invention, the estimated value of the three-dimensional image patch group is obtained by performing an inverse three-dimensional cooperative transformation on each transform component after hard thresholding. : in, This is an estimate of the 3D image patch group obtained by inverse transformation reconstruction of the 3D cooperative transform. Hard threshold filtering for noise reduction; G is the inverse of the three-dimensional cooperative transformation, and G is the transformation component.

[0040] For any pixel in the overlapping region, its denoised pixel value can be obtained by weighting the image block to which it belongs and its weights. : Among them, the initial denoised image is obtained. , Let be the pixel value of the image patch at position (x, y), and the corresponding weight be . .

[0041] The weighted average value of each pixel is obtained by weighted aggregation of each image patch. This is the final image denoising result.

[0042] in, This is the final denoised image obtained by weighted aggregation of each image patch; Let be the pixel value of the image patch at position (x, y), with corresponding weights as follows: Hw .

[0043] like Figure 4 As shown, the bottom echo image processing method provided in this embodiment of the invention specifically includes: (1) Data acquisition and preliminary imaging: The raw bottom echo data acquired by the multibeam sonar is initially unprocessed acoustic echo signal (usually containing information such as water depth and backscatter intensity). Through preliminary data processing and imaging, an initial multibeam bottom echo image is formed. At this time, the image contains obvious intensity anomalies in the central beam region and complex ocean noise.

[0044] (2) Preliminary denoising based on improved Empirical Mode Decomposition (EMD) is performed on the echo signal of each ping (each pulse transmission), decomposing it into a high-frequency IMF and a low-frequency IMF. Wavelet thresholding is used to denoise the high-frequency IMF containing the main noise; the noise components of the low-frequency IMF containing the main signal are zeroed. The processed components are reconstructed to obtain the denoised single ping signal, which solves the problem of complex noise interference and generates a denoised image.

[0045] (3) Image optimization based on improved BM3D: Image after preliminary denoising by EMD (noise is reduced at this point, but central beam anomalies may still exist). Edge detection is performed on the image to extract structural information maps and preserve details in subsequent processing. Similar image blocks are searched in the image to form a three-dimensional array. A three-dimensional transformation (two-dimensional DCT + one-dimensional HT) is performed, and threshold filtering is performed in the transform domain to separate the signal from residual noise. An inverse transformation is performed to obtain the estimated value, and the image is restored by weighted aggregation, which solves the problem of central beam anomalies and further removes residual noise while preserving the edge details of the image.

[0046] (4) After the above two steps of combined processing, a high-quality multibeam bottom echo image is obtained. The image at this time has a high signal-to-noise ratio, uniform intensity, and clear details, and can be directly used for high-precision seabed sediment classification and target detection.

[0047] The bottom echo image processing method provided in this invention analyzes the causes of noise generation in multibeam echo signals, decomposes the signal into high- and low-frequency IMF components using empirical mode decomposition, performs wavelet transform on the high-frequency components, and employs a semi-soft threshold function for filtering. It proposes a threshold setting method based on Stein's Unbiased Risk Estimate (SURE) unbiased estimation, achieving adaptive selection of the filtering threshold. To improve the efficiency and accuracy of similar block matching in the Block-Matching and 3D Filtering (BM3D) algorithm, an edge detection-based similar block search strategy is implemented. Edge information maps are obtained by edge detection of the original image, and a search is performed along the edges during similar block matching, achieving fast and accurate matching of similar blocks. This invention solves the problem of poor quality in multibeam bottom echo images due to complex noise, achieving comprehensive denoising in both the one-dimensional signal domain and the two-dimensional image domain.

[0048] The bottom echo image processing apparatus provided by the present invention is described below. The bottom echo image processing apparatus described below can be referred to in correspondence with the bottom echo image processing method described above.

[0049] Figure 5 A functional structure diagram of the bottom echo image processing device provided in an embodiment of the present invention is shown below. Figure 5 As shown, the bottom echo image processing apparatus provided in this embodiment of the invention includes: The signal domain processing module 501 is used to acquire the original bottom echo signal, perform signal domain processing on the original bottom echo signal, and construct a preliminary bottom echo image. Image domain processing module 502 is used to perform image domain processing on the preliminary bottom echo image, including: extracting the image structure information map of the preliminary bottom echo image; searching for similar image blocks in the preliminary bottom echo image based on the image structure information map; combining the searched similar image blocks into a three-dimensional block group; transforming the three-dimensional group to the transform domain for coefficient shrinkage processing; and generating a bottom echo image through inverse transformation and weighted aggregation.

[0050] The bottom echo image processing apparatus provided in this embodiment of the invention acquires the original bottom echo signal, performs signal domain processing on the original bottom echo signal to construct a preliminary bottom echo image; performs image domain processing on the preliminary bottom echo image, including: extracting the image structure information map of the preliminary bottom echo image; searching for similar image blocks in the preliminary bottom echo image based on the image structure information map; combining the searched similar image blocks into a three-dimensional block group; transforming the three-dimensional group to the transform domain for coefficient shrinkage processing; and generating a bottom echo image through inverse transformation and weighted aggregation. This embodiment of the invention, by adaptively decomposing and differentially denoising the original echo in the signal domain, can effectively suppress various noise interferences introduced by complex marine environments, significantly improving the signal-to-noise ratio and data quality of the bottom echo image; furthermore, by introducing structure-prior-guided similar block matching and three-dimensional collaborative filtering in the image domain, it can accurately preserve the edge details and weak target information of the image while correcting the intensity anomaly in the central beam region, providing reliable data support for high-precision seabed sediment classification and target detection.

[0051] Figure 6 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 6 As shown, the electronic device may include a processor 610, a communication interface 620, a memory 630, and a communication bus 640. The processor 610, communication interface 620, and memory 630 communicate with each other via the communication bus 640. The memory 630 includes computer programs, an operating system, and acquired data. The processor 610 can call logical instructions in the memory 630 to execute a bottom echo image processing method. This method includes: acquiring the original bottom echo signal; performing signal domain processing on the original bottom echo signal to construct a preliminary bottom echo image; performing image domain processing on the preliminary bottom echo image, including: extracting an image structure information map from the preliminary bottom echo image; searching for similar image blocks in the preliminary bottom echo image based on the image structure information map; combining the searched similar image blocks into a three-dimensional block group; transforming the three-dimensional group to the transform domain for coefficient shrinkage processing; and generating a bottom echo image through inverse transform and weighted aggregation.

[0052] Furthermore, the logical instructions in the aforementioned memory 630 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to related technologies, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0053] On the other hand, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the bottom echo image processing method provided by the above methods. The method includes: acquiring an original bottom echo signal; performing signal domain processing on the original bottom echo signal to construct a preliminary bottom echo image; performing image domain processing on the preliminary bottom echo image, including: extracting an image structure information map of the preliminary bottom echo image; searching for similar image blocks in the preliminary bottom echo image based on the image structure information map; and combining the searched similar image blocks into a three-dimensional block group; transforming the three-dimensional group to the transform domain for coefficient shrinkage processing; and generating a bottom echo image through inverse transformation and weighted aggregation.

[0054] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0055] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the parts that contribute to the related technology, can be embodied in the form of software products. This computer software product can be stored in a computer-readable 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 the various embodiments or some parts of the embodiments.

[0056] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for processing bottom echo images, characterized in that, include: The raw bottom echo signal is acquired, and the raw bottom echo signal is processed in the signal domain to construct a preliminary bottom echo image. Image domain processing of the preliminary bottom echo image includes: extracting the image structure information map of the preliminary bottom echo image; searching for similar image blocks in the preliminary bottom echo image based on the image structure information map; combining the searched similar image blocks into a three-dimensional block group; transforming the three-dimensional group to the transform domain for coefficient shrinkage processing; and generating a bottom echo image through inverse transformation and weighted aggregation.

2. The bottom echo image processing method according to claim 1, characterized in that, The step of performing signal domain processing on the original bottom echo signal to construct a preliminary bottom echo image includes: Empirical mode decomposition is performed on the original bottom echo signal to obtain several intrinsic mode function components and residual components; Calculate the correlation coefficient between each intrinsic mode function component and the original bottom echo signal, and divide each intrinsic mode function component into high-frequency components and low-frequency components based on the correlation coefficient; The high-frequency components are subjected to wavelet threshold denoising, while the low-frequency components and residual components are subjected to noise component zeroing. The high-frequency components that have undergone noise reduction, the low-frequency components that have been zeroed out, and the residual components are reconstructed to obtain the noise-reduced low echo signal. The noise-reduced bottom echo signals corresponding to each beam position are superimposed to generate a preliminary bottom echo image.

3. The bottom echo image processing method according to claim 1, characterized in that, The extraction of the image structure information map from the preliminary bottom echo image includes: Edge detection is performed on the preliminary bottom echo image, and the edge detection results are used as the image structure information map.

4. The bottom echo image processing method according to claim 1 or 3, characterized in that, The step of searching for similar image patches in the preliminary bottom echo image based on the image structure information map, and combining the searched similar image patches into a three-dimensional block group, includes: A reference image block is selected from the preliminary bottom echo image; Based on the image structure information map, the similarity metric between the candidate image block and the reference image block is calculated within the search window, and the candidate image block that meets the preset similarity condition is taken as a similar image block. The reference image block and its corresponding similar image block are stacked along the third dimension to form a three-dimensional block group.

5. The bottom echo image processing method according to claim 1, characterized in that, The step of transforming the three-dimensional group to the transform domain and performing coefficient shrinkage includes: A two-dimensional transformation is performed on each image block in the three-dimensional block group, and a one-dimensional transformation is performed on the stacking dimension of the three-dimensional block group to obtain the three-dimensional transformation coefficients. Threshold filtering is applied to the three-dimensional transformation coefficients, retaining those greater than a preset threshold and setting coefficients not greater than the preset threshold to zero; The preset threshold is determined based on the noise level estimate of the high-frequency components in the transform domain.

6. The bottom echo image processing method according to claim 5, characterized in that, The inverse transform includes: The three-dimensional transformation coefficients that are greater than the preset threshold are subjected to a one-dimensional inverse transformation along the stacking dimension, and a two-dimensional inverse transformation is performed on each image block to obtain the denoised three-dimensional block group estimate.

7. The bottom echo image processing method according to claim 6, characterized in that, The weighted aggregation includes: Determine the weight of each pixel within each image block in each of the three-dimensional block group estimates; For each pixel in the preliminary bottom echo image, the estimated values ​​of all image blocks containing that pixel at that pixel and their corresponding weights are obtained, and a weighted average is performed to obtain the final pixel value of that pixel. A bottom echo image is generated based on the final pixel values ​​of all pixels.

8. A bottom echo image processing device, characterized in that, include: The signal domain processing module is used to acquire the original bottom echo signal, perform signal domain processing on the original bottom echo signal, and construct a preliminary bottom echo image. The image domain processing module is used to perform image domain processing on the preliminary bottom echo image, including: extracting the image structure information map of the preliminary bottom echo image; searching for similar image blocks in the preliminary bottom echo image based on the image structure information map; combining the searched similar image blocks into a three-dimensional block group; transforming the three-dimensional group to the transform domain for coefficient shrinkage processing; and generating a bottom echo image through inverse transformation and weighted aggregation.

9. 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 program, it implements the bottom echo image processing method as described in any one of claims 1 to 7.

10. A non-transitory readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the bottom echo image processing method as described in any one of claims 1 to 7.

Images