A 3D reconstruction method of side scan sonar images based on SFS technology

Abstract

The present invention discloses a three-dimensional reconstruction method of side-scan sonar images based on SFS technology, which is applicable to the field of ocean detection and survey. The method comprises the following steps: S1, seabed line detection. The seabed line detection method based on the Canny operator extracts the seabed contour line in the sonar image to ensure the accuracy of subsequent processing; S2, grayscale correction. First, the image is divided into four sub-image areas with equal widths, and then a six-segment gain compensation curve is generated according to the mean of each sub-image area. Finally, the side-scan sonar image is gain compensated according to the gain compensation equation to correct the illumination and contrast changes in the image and improve the image quality; S3, slant range correction. The image is geometrically corrected to eliminate the distance error caused by the tilt of the sonar equipment and accurately characterize the actual shape and size of the measured seabed target; S4, three-dimensional reconstruction of SFS side-scan sonar images based on wavelet transform. By utilizing the multi-resolution characteristics of wavelets, images can be decomposed and reconstructed. First, the image is decomposed into rows to obtain the low-frequency component L and high-frequency component H of the image in the horizontal direction. Then, column decomposition is performed on this basis to finally obtain a low-frequency approximate coefficient LL and three high-frequency component images of horizontal details LH, vertical details HL, and diagonal details HH. When reconstructing the image, first perform a one-dimensional discrete wavelet inverse transform on each column of the decomposition result, and then perform a one-dimensional discrete wavelet inverse transform on each row of the obtained data to obtain the reconstructed image.

Description

Technical Field

[0001] The invention belongs to image processing technology, relates to the field of ocean detection and survey technology, and particularly to a three-dimensional reconstruction method of side-scan sonar images based on SFS technology. Background Art

[0002] Seabed topography detection is a complex process involving a variety of technologies and equipment, the purpose of which is to obtain detailed information on seabed topography, geological structure and other features. This is of great significance to scientific research, marine engineering, environmental protection and resource development. With the rapid development of sonar technology, side-scan sonar systems are increasingly used in marine development. However, due to the measurement mechanism and operation mode limitations of side-scan sonar, it cannot directly provide information on the seabed surface topography and the height of underwater targets. Therefore, the study of three-dimensional reconstruction of the original two-dimensional grayscale images of side-scan sonar is particularly important.

[0003] Traditional 3D reconstruction methods face limitations in accuracy and resolution, severe noise interference, slow processing speed and low efficiency, and poor adaptability in extreme or complex environments. Compared with traditional methods, the 3D reconstruction method of side scan sonar images based on SFS technology has significant advantages. First, it can improve the accuracy and resolution of 3D reconstruction, overcoming the limitations of traditional methods in noise and resolution. Second, SFS technology enhances the ability to restore details, making the reconstruction of small features and complex structures of seabed terrain and underwater targets more accurate. In addition, it improves image fusion and processing, and can better synthesize data from different angles or time points to provide a comprehensive and consistent 3D model. The method based on SFS technology also has high automation and processing efficiency, reduces dependence on human intervention, and shows strong robustness to complex environments, and can provide stable 3D reconstruction results under various marine conditions. Finally, this technology supports real-time and online processing, and can provide terrain and target information in real time during data acquisition, which is of great significance for real-time monitoring and decision support. Summary of the invention

[0004] In order to solve the above problems in the prior art, the present invention provides a three-dimensional reconstruction method of side-scan sonar images based on SFS technology.

[0005] To achieve the above object, the present invention provides the following technical solution: a three-dimensional reconstruction method of side scan sonar images based on SFS technology, the method comprising the following steps:

[0006] S1. Seabed line detection. The seabed line detection method based on the Canny operator extracts the seabed contour line in the sonar image to ensure the accuracy of subsequent processing;

[0007] S2. Grayscale correction. First, the image is divided into four sub-image areas with equal widths. Then, a six-segment gain compensation curve is generated according to the mean value of each sub-image area. Finally, the side scan sonar image is gain compensated according to the gain compensation equation to correct the illumination and contrast changes in the image and improve the image quality.

[0008] S3, Slant range correction. Perform geometric correction on the image to eliminate the distance error caused by the tilt of the sonar equipment and accurately characterize the actual shape and size of the measured seabed target;

[0009] S4. Three-dimensional reconstruction of SFS side scan sonar images based on wavelet transform. The multi-resolution characteristics of wavelets can be used to decompose and reconstruct images. First, the image is decomposed into rows to obtain the low-frequency component L and high-frequency component H of the image in the horizontal direction. Then, column decomposition is performed on this basis to finally obtain a low-frequency approximate coefficient LL and three high-frequency component images of horizontal details LH, vertical details HL and diagonal details HH. When reconstructing the image, first perform a one-dimensional discrete wavelet inverse transform on each column of the decomposition result, and then perform a one-dimensional discrete wavelet inverse transform on each row of the obtained data to obtain the reconstructed image.

[0010] Preferably, the specific content of S1 is:

[0011] Step 1: Use a Gaussian filter to smooth the sonar image to reduce the impact of noise.

[0012] Step 2: Apply the Canny edge detection operator and use the gradient operator to calculate the gradient strength and direction of each pixel in the image to identify potential edges; non-maximum suppression removes non-edge pixels, and double threshold processing divides the edges into strong edges and weak edges, and completes the final edge detection through edge connection.

[0013] Step 3: Extract the seabed line, extract the seabed contour line from the Canny edge detection results.

[0014] Preferably, the specific content of S2 is:

[0015] Step 1: Divide the image into four sub-image areas with equal widths and calculate the regional average values ​​M1, M2, M3, M4 and the overall mean M of the four windows respectively.

[0016] Step 2: Calculate the compensation factor P i :

[0017]

[0018] Among them, M represents the overall mean, M i Represents the average of four sub-image regions with equal width.

[0019] Step 3: Obtain the compensation equation. The calculation formula of the compensation equation is:

[0020]

[0021] Where p1, p2, p3 and p4 are compensation factors and W is the image width.

[0022] Step 4: Based on the side scan sonar image compensation result obtained from the compensation curve, correct the illumination and contrast changes in the image to improve the image quality.

[0023] Preferably, the specific content of S3 is:

[0024] Assuming the seabed surface is horizontal, the horizontal distance of the seabed echo point is:

[0025]

[0026] Among them, HD represents horizontal distance, SD represents slant distance, and H represents the vertical distance between the sonar and the seabed;

[0027] The image is geometrically corrected to eliminate the distance error caused by the tilt of the sonar equipment and accurately characterize the actual shape and size of the measured seabed target.

[0028] Preferably, the specific content of S4 is:

[0029] Step 1: Use wavelet db2 to decompose the two-dimensional side-scan sonar waterfall image to obtain a low-frequency component LL and three high-frequency components: horizontal details LH, vertical details HL and diagonal details HH.

[0030] Step 2: A gradient algorithm based on the Priwitt operator is used in the low-frequency component of the image to directly obtain gradient information from the image and calculate the three-dimensional height value of the seabed surface in the low-frequency part.

[0031] Step 3: Use the classical minimization method to iteratively solve the three high-frequency components of the image and calculate the height values ​​of the corresponding points.

[0032] Step 4: Combine the experimental results obtained in the second and third steps to achieve three-dimensional reconstruction of the side-scan sonar image and restore the three-dimensional topography of the seabed surface.

[0033] Implementing the embodiments of the present invention will have the following beneficial effects:

[0034] The 3D reconstruction method of side-scan sonar images based on SFS technology can significantly improve the reconstruction accuracy, accurately depict the seabed topography, and overcome the accuracy limitations of traditional technologies. Using wavelet transform and optimization algorithms, the processing speed is significantly improved, especially when processing large-scale data. At the same time, the high degree of automation of the method reduces manual intervention, improves processing efficiency and consistency of results. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the embodiments of the present invention or the technical solutions of the prior art, the relevant drawings are briefly introduced. These drawings only represent some embodiments of the present invention and provide technical professionals with the possibility of obtaining other information based on these drawings without performing creative work. The drawings intuitively present the structure, components or processes of the present invention to better understand the details of the embodiments. In general, these drawings are intended to enable professionals in the technical field to understand the embodiments of the present invention and provide them with sufficient information to design and implement similar or equivalent drawings. Among them:

[0036] Figure 1 It is a flow chart of a three-dimensional reconstruction method of side scan sonar images based on SFS technology;

[0037] Figure 2 A side scan sonar image in one embodiment;

[0038] Figure 3 is a side scan sonar image preprocessing result described in one embodiment;

[0039] Figure 4 3D reconstruction according to an embodiment. DETAILED DESCRIPTION

[0040] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0041] Implementing the embodiments of the present invention will have the following beneficial effects:

[0042] In order to solve the shortcomings of the existing technology, in this embodiment, a side scan sonar image 3D reconstruction method based on SFS technology is proposed. Figures 1 to 4 As shown, the method includes:

[0043] S1. Seabed line detection. The seabed line detection method based on the Canny operator extracts the seabed contour line in the sonar image to ensure the accuracy of subsequent processing; S2. Grayscale correction. First, the image is divided into four sub-image areas with equal width, and then a six-segment gain compensation curve is generated according to the mean of each sub-image area. Finally, the side scan sonar image is gain compensated according to the gain compensation equation to correct the illumination and contrast changes in the image and improve the image quality; S3. Slant range correction. The image is geometrically corrected to eliminate the distance error caused by the tilt of the sonar equipment and accurately characterize the actual shape and size of the measured seabed target; S4. Three-dimensional reconstruction of SFS side scan sonar images based on wavelet transform. Using the multi-resolution characteristics of wavelets, the image can be decomposed and reconstructed. First, the image is decomposed into rows to obtain the low-frequency component L and high-frequency component H of the image in the horizontal direction, and then column decomposition is performed on this basis to finally obtain a low-frequency approximation coefficient LL and three high-frequency component images of horizontal detail LH, vertical detail HL and diagonal detail HH. When reconstructing an image, first perform a one-dimensional discrete wavelet inverse transform on each column of the decomposition result, and then perform a one-dimensional discrete wavelet inverse transform on each row of the obtained data to obtain the reconstructed image.

[0044] In some specific embodiments, the specific content of S1 is:

[0045] Step 1: Use a Gaussian filter to filter the side scan sonar image Figure 1 Perform smoothing preprocessing to reduce the impact of noise;

[0046] Step 2: Apply the Canny edge detection operator and use the gradient operator to calculate the gradient strength and direction of each pixel in the image to identify potential edges; non-maximum suppression removes non-edge pixels, and double threshold processing divides the edges into strong edges and weak edges, and completes the final edge detection through edge connection;

[0047] Step 3: Extract the seabed line, extract the seabed contour line from the Canny edge detection results.

[0048] In some specific embodiments, the specific content of S2 is:

[0049] Step 1: Divide the image into four sub-image areas with equal widths and calculate the regional average values ​​M1, M2, M3, M4 and the overall mean M of the four windows respectively;

[0050] Step 2: Calculate the compensation factor P i :

[0051]

[0052] Among them, M represents the overall mean, M irepresents the average value of four sub-image regions with equal width;

[0053] Step 3: Obtain the compensation equation. The calculation formula of the compensation equation is:

[0054]

[0055] Where p1, p2, p3 and p4 are compensation factors, and W is the image width;

[0056] Step 4: Based on the side scan sonar image compensation result obtained from the compensation curve, correct the illumination and contrast changes in the image to improve the image quality.

[0057] In some specific embodiments, the specific content of S3 is:

[0058] Assuming the seabed surface is horizontal, the horizontal distance of the seabed echo point is:

[0059]

[0060] Among them, HD represents horizontal distance, SD represents slant distance, and H represents the vertical distance between the sonar and the seabed;

[0061] The image is geometrically corrected to eliminate the distance error caused by the tilt of the sonar equipment and accurately characterize the actual shape and size of the measured seabed target. Figure 3 It is the result of image preprocessing.

[0062] In some specific embodiments, the specific content of S4 is:

[0063] Step 1: Use wavelet db2 to decompose the two-dimensional side-scan sonar waterfall image to obtain a low-frequency component LL and three high-frequency components: horizontal details LH, vertical details HL, and diagonal details HH;

[0064] Step 2: A gradient algorithm based on the Priwitt operator is used in the low-frequency component of the image to directly obtain gradient information from the image and calculate the three-dimensional height value of the seabed surface in the low-frequency component.

[0065] Step 3: Use the classical minimization method to iteratively solve the three high-frequency components of the image and calculate the height values ​​of the corresponding points;

[0066] Step 4: Combine the experimental results obtained in steps 2 and 3 to achieve 3D reconstruction of the side-scan sonar image and restore the 3D topography of the seabed surface. Figure 4 is the three-dimensional reconstruction image.

[0067] Aiming at the problems of insufficient accuracy and low processing efficiency of existing side scan sonar image 3D reconstruction, the present invention proposes a side scan sonar image 3D reconstruction method based on SFS technology. This method significantly improves the image reconstruction accuracy and quality, reduces noise interference, and optimizes the processing speed through the steps of seabed line detection, grayscale correction, slant range correction, and 3D reconstruction based on wavelet transform.

[0068] In summary, this side-scan sonar image 3D reconstruction method based on SFS technology has significant advantages in providing high-precision 3D reconstruction results, improving image quality and processing efficiency, and can effectively meet application requirements in complex environments.

[0069] The above description only represents a preferred implementation in the field of seabed surveying, but is not limited to this. The protection scope of the present invention covers a wider range of implementation forms. The technical solution mentioned is only a specific implementation based on the concept of the present invention, not the only form. It should be emphasized that the protection scope of the present invention is far more than the above examples. Under the basic concept of the present invention, various improvements, changes, applications and embodiments are within its protection scope. For technical experts in this field, as long as they do not deviate from the core principles of the present invention, any modification or change is deemed to be within the protection scope of the present invention. The protection scope of the present invention includes a variety of possible implementation methods and technical solutions, allowing technicians to make appropriate adjustments and innovations according to actual needs without violating the core idea of ​​the present invention.

Claims

1. A three-dimensional reconstruction method of side scan sonar images based on SFS technology, characterized in that It includes the following steps: S1. Seabed line detection. The seabed line detection method based on the Canny operator extracts the seabed contour line in the sonar image to ensure the accuracy of subsequent processing; S2. Grayscale correction. First, the image is divided into four sub-image areas with equal widths. Then, a six-segment gain compensation curve is generated according to the mean value of each sub-image area. Finally, the side scan sonar image is gain compensated according to the gain compensation equation to correct the illumination and contrast changes in the image and improve the image quality. S3, Slant range correction. Perform geometric correction on the image to eliminate the distance error caused by the tilt of the sonar equipment and accurately characterize the actual shape and size of the measured seabed target; S4. Three-dimensional reconstruction of SFS side scan sonar images based on wavelet transform. The multi-resolution characteristics of wavelets can be used to decompose and reconstruct images. First, the image is decomposed into rows to obtain the low-frequency component L and high-frequency component H of the image in the horizontal direction. Then, column decomposition is performed on this basis to finally obtain a low-frequency approximate coefficient LL and three high-frequency component images of horizontal details LH, vertical details HL and diagonal details HH. When reconstructing the image, first perform a one-dimensional discrete wavelet inverse transform on each column of the decomposition result, and then perform a one-dimensional discrete wavelet inverse transform on each row of the obtained data to obtain the reconstructed image.

2. The method for three-dimensional reconstruction of side scan sonar images based on SFS technology according to claim 1, characterized in that: The specific content of S1 is: Step 1: Use Gaussian filter to smooth the sonar image to reduce the impact of noise; Step 2: Apply the Canny edge detection operator and use the gradient operator to calculate the gradient strength and direction of each pixel in the image to identify potential edges; non-maximum suppression removes non-edge pixels, and double threshold processing divides the edges into strong edges and weak edges, and completes the final edge detection through edge connection; Step 3: Extract the seabed line, extract the seabed contour line from the Canny edge detection results.

3. The method for three-dimensional reconstruction of side scan sonar images based on SFS technology according to claim 1 is characterized by: The specific content of S2 is: Step 1: Divide the image into four sub-image areas with equal widths and calculate the regional average values ​​M1, M2, M3, M4 and the overall mean M of the four windows respectively; Step 2: Calculate the compensation factor P i : Among them, M represents the overall mean, M i represents the average value of four sub-image regions with equal width; Step 3: Obtain the compensation equation. The calculation formula of the compensation equation is: Where p1, p2, p3 and p4 are compensation factors, and W is the image width; Step 4: Based on the side scan sonar image compensation result obtained from the compensation curve, correct the illumination and contrast changes in the image to improve the image quality.

4. The method for three-dimensional reconstruction of side scan sonar images based on SFS technology according to claim 1 is characterized by: The specific content of S3 is: Assuming the seabed surface is horizontal, the horizontal distance of the seabed echo point is: Among them, HD represents horizontal distance, SD represents slant distance, and H represents the vertical distance between the sonar and the seabed; The image is geometrically corrected to eliminate the distance error caused by the tilt of the sonar equipment and accurately characterize the actual shape and size of the measured seabed target.

5. The method for three-dimensional reconstruction of side scan sonar images based on SFS technology according to claim 1, characterized in that: The specific content of S4 is: Step 1: Use wavelet db2 to decompose the two-dimensional side-scan sonar waterfall image to obtain a low-frequency component LL and three high-frequency components: horizontal details LH, vertical details HL, and diagonal details HH; Step 2: A gradient algorithm based on the Priwitt operator is used in the low-frequency component of the image to directly obtain gradient information from the image and calculate the three-dimensional height value of the seabed surface in the low-frequency component. Step 3: Use the classical minimization method to iteratively solve the three high-frequency components of the image and calculate the height values ​​of the corresponding points; Step 4: Combine the experimental results obtained in the second and third steps to achieve three-dimensional reconstruction of the side-scan sonar image and restore the three-dimensional topography of the seabed surface.

Images