Three dimensional reconstruction of side-scan sonar images

Abstract

To improve restoration quality of three dimensional information of a side scan sonar image, and to improve accuracy of three dimensional reconfiguration of the side scan sonar image.SOLUTION: The three dimensional reconstruction method for side-scan sonar images of the present invention comprises: reading side-scan sonar images to be three dimensionally reconstructed; designing a shape recovery method (SFS) mixture reflection model for side-scan sonar images, including designing an adaptive specular reflection coefficient, designing an adaptive diffuse reflection coefficient, designing an adaptive specular reflection index, and designing a seabed surface adaptive mixture reflection model formula; obtaining three dimensionally reconstructed surface elevations, including inversely transforming seabed surface elevation values and designing an inversely transformed elevation limiting factor; and outputting a three dimensional reconstruction result of the side-scan sonar images. To provide high-quality underwater three dimensional topographical information by application scenes such as AUV underwater high-precision navigation, underwater rescue, and submarine exploration.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to the field of sonar image processing, and in particular to a method for three-dimensional reconstruction of side-scan sonar images. [Background technology]

[0002] Side-scan sonar is widely used in seafloor exploration, underwater rescue, and marine surveys because it can economically and efficiently acquire high-resolution topographical images of the seafloor. However, side-scan sonar is limited by its imaging mechanism and cannot directly capture three-dimensional information about the seafloor. Furthermore, it is difficult to directly represent the undulating features of the seafloor. Therefore, it is necessary to restore the three-dimensional information of the seafloor by three-dimensional reconstruction. Therefore, how to reconstruct three-dimensional information from side-scan sonar images has already become a research topic in the field of underwater acoustics.

[0003] Conventional 3D reconstruction methods for side-scan sonar images can be divided into two categories: shading length reconstruction and shape from shading (SFS). The former is limited to single-target reconstruction and cannot reconstruct large-scale topography. The latter primarily utilizes shading variations in side-scan images to reconstruct 3D seafloor topography. It is suitable for large-area reconstruction. Therefore, research into 3D reconstruction based on SFS is underway. The reflectivity model in SFS is the key to 3D reconstruction, and adapting the reflectivity model to the seafloor surface features has become a key factor in 3D topography reconstruction. Therefore, research into 3D reconstruction methods for side-scan sonar images will help to accurately reconstruct the stereoscopic information of side-scan sonar images and improve the accuracy of 3D reconstruction of side-scan sonar images in complex seafloor environments. Summary of the Invention [Problem to be solved by the invention]

[0004] The present invention aims to provide a method for three-dimensional reconstruction of side scan sonar images that improves the quality of restoration of three-dimensional information from side scan sonar images and further improves the accuracy of three-dimensional reconstruction of side scan sonar images. [Means for solving the problem]

[0005] The three-dimensional reconstruction method for side-scan sonar images first reads the side-scan sonar image to be three-dimensionally reconstructed, then designs the SFS mixed reflection model for the side-scan sonar image, then calculates the three-dimensional reconstruction surface elevation, and finally outputs the three-dimensional reconstruction result of the side-scan sonar image.

[0006] in particular, Step 1: reading a side-scan sonar image of a target for three-dimensional reconstruction; Adaptive specular reflection coefficient D g and adaptive diffuse reflection coefficient D r Step 2 of designing an SFS mixed reflection model for side-scan sonar images, including designing an adaptive specular reflectance exponent ε and designing an adaptive seabed surface mixed reflection model formula; The seabed surface elevation value is inversely converted, and the inverse conversion elevation control coefficient k i Step 3 of determining the 3D reconstructed surface elevation, which includes designing Step 4 is to output the three-dimensional reconstruction result of the side scan sonar image.

[0007] Furthermore, designing the SFS mixed reflection model for the side scan sonar image in step 2 includes: g and adaptive diffuse reflection coefficient D r designing an adaptive specular reflection exponent ε; and designing an adaptive mixed reflection model formula for the seabed surface; g Specifically, designing the system involves steps 1) and 2), 1) Calculate the surface roughness C of the side scan sonar image, 2) Adaptive specular reflection coefficient D gSpecifically, however, we adopt equation (1).

number

[0008] the adaptive diffuse reflection coefficient D r Specifically, the design employs Eq. (2).

number

[0009] Specifically, designing the adaptive specular reflection index ε employs Equation (3):

number

[0010] The design of the seabed surface adaptive mixed reflection model formula specifically employs Equation (4):

number

[0011] Furthermore, in designing the seabed surface adaptive mixed reflection model formula described in step 2, the parameters φ and τ are specifically adopted in steps a) to c), a) Set the coordinate system y-axis of a point on the seabed surface so that the direction of the side scan sonar track line is the same. At this time, the incident angle τ of the side scan sonar signal is 0. b) Based on the working mechanism of the side scan sonar, the tilt angle φ(x,i) corresponding to the x-th pixel of the i-th frame of the side scan sonar image is given by Equation (5):

number

[0012] Furthermore, the inverse transformation of the seabed surface elevation value in step (3) specifically employs steps (I) to (V), (I) The seabed surface slopes p and q are approximated using the backward finite difference method and discretized as in equations (6) and (7).

number

number

number

number

[0013] Furthermore, the inverse transformation altitude restriction coefficient k i Specifically, the design involves steps (A) and (B), (A) The seafloor point position in the ith frame of the side scan sonar image is W i If we set it as follows, the current depth H i is Equation (13),

number

number

[0014] Furthermore, the step 3 of obtaining the three-dimensional reconstructed surface elevation specifically includes steps (a) to (e), (a) The seafloor surface elevation values ​​of the side scan sonar image are inversely transformed, (b) When all depth data on the track is obtained, the depth set H = {H1, H2..., H i …,H o}, (c) Inverse transformation altitude control coefficient k of the ith frame of the side scan sonar image i Calculate (d) The equation for the inverse transformation elevation value Φ(Z(x,i)) corresponding to the i-th frame of the regulated side-scan sonar image is Equation (15),

number

[0015] Compared with the prior art, the method of the present invention involves first reading a side-scan sonar image of a three-dimensional reconstruction target, and then calculating an adaptive specular reflectance coefficient D g and adaptive diffuse reflection coefficient D r , designing an adaptive specular reflectance exponent ε, and designing an adaptive seabed surface mixed reflection model formula; and further inversely transforming the seabed surface elevation value and determining an inverse transformation elevation control coefficient k i and finally, a step of outputting the 3D reconstruction result of the side scan sonar image. Through the above steps, the 3D reconstruction method of side scan sonar images of the present invention is useful for restoring the stereoscopic information of side scan sonar images with high accuracy, improving the accuracy of 3D reconstruction of side scan sonar images in complex seabed environments, providing more intuitive and reliable underwater 3D terrain information, and providing higher quality underwater 3D terrain information for applications such as AUV underwater high-precision navigation, underwater rescue, and seabed exploration. Specifically, it has the following advantages:

[0016] (1) The SFS mixed reflection model of side-scan sonar images designed in this invention is based on the fact that the seafloor surface of the continental shelf is relatively flat and has characteristics such as the presence of reefs, and includes not only diffuse reflection but also specular reflection. While retaining the Lambert diffuse reflection model, this invention introduces the Blinn-Phong specular reflection model to form an SFS mixed reflection model that adapts to complex seafloor surfaces, thereby enhancing the matching between the reflection model and the seafloor surface and thereby improving the accuracy of three-dimensional reconstruction of side-scan sonar images.

[0017] (2) The adaptive mixed reflection model equation for the seabed surface designed in this invention is an exponential function with an adaptive specular reflection coefficient D g and D using a fractional formula including the roughness C of the side scan sonar image. g The basis number of is expressed as an index, the roughness C is used as an index, the relationship between the side scan sonar image and the seafloor surface reflection model is improved, an adaptive specular reflectance index ε is designed, and D g is used as a coefficient in the formula for the specular reflection index ε, and ε is expressed by multiplying the inverse of the roughness by a constant λ, and the adaptive diffuse reflection coefficient D r By designing a new SFS mixed reflection model, we can improve the accuracy of three-dimensional reconstruction of side-scan sonar images.

[0018] (3) The inverse transformation altitude restriction coefficient k designed in this invention i The depth data set H = {H1, H2..., Hi..., Ho} on the track line is calculated using the seafloor point positions of the side scan sonar image, and the inverse transformation elevation value of the depth data and the track line position is calculated.

number

[0019] [Figure 1] 1 is an overall flowchart of a method for three-dimensional reconstruction of a side-scan sonar image according to the present invention. [Figure 2] This is a distribution characteristic map of the seabed surface. [Figure 3] Figure 3 is a comparison diagram of three-dimensional reconstructions of side-scan sonar image I, where Figure 3(a) shows side-scan sonar image I, Figure 3(b) shows a three-dimensional reconstruction using the Lambert model method, Figure 3(c) shows a three-dimensional reconstruction using the Ward mixture model method, and Figure 3(d) shows a three-dimensional reconstruction using the method of the present invention. [Figure 4] Figure 4 is a comparison diagram of three-dimensional reconstructions of side scan sonar image II, where Figure 4(a) shows side scan sonar image II, Figure 4(b) shows a three-dimensional reconstruction diagram using the Lambert model method, Figure 4(c) shows a three-dimensional reconstruction diagram using the Ward mixture model method, and Figure 4(d) shows a three-dimensional reconstruction diagram using the method of the present invention. [Figure 5] Figure 5 is a comparison diagram of three-dimensional reconstructions of side scan sonar image III, where Figure 5(a) shows side scan sonar image III, Figure 5(b) shows a three-dimensional reconstruction diagram using the Lambert model method, Figure 5(c) shows a three-dimensional reconstruction diagram using the Ward mixture model method, and Figure 5(d) shows a three-dimensional reconstruction diagram using the method of the present invention. [Figure 6] Figure 6 is a comparison diagram of three-dimensional reconstructions of side-scan sonar image IV, where Figure 6(a) shows side-scan sonar image IV, Figure 6(b) shows a three-dimensional reconstruction using the Lambert model method, Figure 6(c) shows a three-dimensional reconstruction using the Ward mixture model method, and Figure 6(d) shows a three-dimensional reconstruction using the method of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0020] In order to clarify the objectives and technical solutions of the embodiments of the present invention, the technical solutions of the present invention will be further described with reference to the drawings. Obviously, the described embodiments are only some of the embodiments of the present invention, but not all of the embodiments. Based on the described embodiments of the present invention, those skilled in the art can easily obtain other embodiments without any creative efforts, and all other embodiments also fall within the scope of protection of the present invention.

[0021] As shown in Figure 1, the method for 3D reconstruction of side-scan sonar images of the present invention improves the quality of restoration of the three-dimensional information of side-scan sonar images and further improves the accuracy of 3D reconstruction of side-scan sonar images. First, the side-scan sonar image to be 3D reconstructed is read, then an SFS mixed reflection model of the side-scan sonar image is designed, then the 3D reconstruction surface elevation is calculated, and finally the 3D reconstruction result of the side-scan sonar image is output. Specifically, the following steps 1 to 4 are adopted.

[0022] Step 1: Read the side-scan sonar image of the target for three-dimensional reconstruction.

[0023] Step 2: Design the SFS mixed reflection model for side-scan sonar images.

[0024] Designing the SFS mixed reflection model for the side scan sonar image includes adaptive specular reflectance coefficients D g and adaptive diffuse reflection coefficient D rThis involves designing an adaptive specular reflectance exponent ε and an adaptive mixed reflection model for the seafloor surface. Side-scan sonar can capture high-resolution two-dimensional seafloor images, but it cannot intuitively represent seafloor topography. Related research has shown that the echo intensity of side-scan sonar is related to the undulations of the seafloor and can reflect relevant information about the seafloor. Based on this, mining more seafloor topography information can provide a basis for seafloor topography reconstruction. To address the drawback of side-scan sonar images being unable to intuitively represent seafloor topography, a shape-from-shading (SFS) method is used to reconstruct three-dimensional side-scan sonar images. As shown in Figure 2, the seafloor surface of the continental shelf is relatively flat and characterized by the presence of reefs, resulting in both diffuse and specular reflection. Therefore, this invention retains the Lambertian model and introduces the Blinn-Phong model to represent the specular reflection of the seafloor.

[0025] the adaptive specular reflection coefficient D g Specifically, designing the system involves steps 1) and 2), 1) Calculate the surface roughness C of the side scan sonar image, 2) Adaptive specular reflection coefficient D g Specifically, Equation (1) is used.

number

[0026] the adaptive diffuse reflection coefficient D r Specifically, the design employs Eq. (2).

number

[0027] Specifically, designing the adaptive specular reflection index ε employs Equation (3):

number

[0028] The design of the seabed surface adaptive mixed reflection model formula specifically includes steps a) to c), a) The seabed surface adaptive mixed reflection model formula is designed as shown in Equation (4),

number

number

[0029] Step 3: Obtain the 3D reconstructed surface elevation.

[0030] The three-dimensional reconstructed surface elevation is calculated by inversely transforming the seabed surface elevation value and calculating the inverse transformation elevation control coefficient k iThis involves designing a linearization method for the SFS 3D reconstruction of side-scan sonar images. This involves first approximating the partial differentials of the irradiance equation using finite difference operations, then discretizing the irradiance equation, and then using the Newton-Raphson method to obtain the 3D reconstructed surface elevation. Linearization of the SFS 3D reconstruction of side-scan sonar images, i.e., linearizing the complex seafloor surface reflection model function of this invention, converts the original nonlinear SFS problem into a linear problem for solution.

[0031] Specifically, the inverse transformation of the seabed surface elevation value includes steps (I) to (V), (I) The seabed surface slopes p and q are approximated using the backward finite difference method and discretized as in equations (6) and (7).

number

number

number

number

[0032] Current 3D reconstruction methods for SFS side-scan sonar images can only capture the relative topography of the seafloor, not the absolute topography. Therefore, researchers rely on external multi-beam bathymetric data to determine the 3D reconstruction. External bathymetric data requires the use of a multi-beam bathymetric sonar to measure the same area of ​​the seafloor, which reduces the efficiency of 3D reconstruction to some extent, and excessive reliance on external bathymetric data reduces the effectiveness of the reflection model. Therefore, the present invention does not rely on external bathymetric data to determine the inversion topography, but instead uses seafloor point data extracted from side-scan sonar images to determine the inversion topography. The seafloor points correspond to the depth values ​​from the current side-scan sonar away from the seafloor. After obtaining all the seafloor point locations, the depth data of the AUV's track can be obtained.

[0033] The inverse transformation altitude restriction coefficient k i Specifically, the design involves steps (A) and (B), (A) The seafloor point position in the ith frame of the side scan sonar image is W i If we set it as follows, the current depth H i is Equation (13),

number

number

[0034] Specifically, the three-dimensional reconstructed surface elevation is calculated by employing steps (a) to (e), (a) The seafloor surface elevation values ​​of the side scan sonar image are inversely transformed, (b) When all depth data on the track is obtained, the depth set H = {H1, H2..., H i …,H o}, (c) Inverse transformation altitude control coefficient k of the ith frame of the side scan sonar image i Calculate (d) The equation for the inverse transformation elevation value Φ(Z(x,i)) corresponding to the i-th frame of the regulated side-scan sonar image is Equation (15),

number

[0035] Step 4: Output the 3D reconstruction results of the side-scan sonar image.

[0036] To verify the effectiveness of the 3D reconstruction method of side-scan sonar images of the present invention, we selected side-scan sonar images of the seabed topography and the underwater wind power generation pile foundation to verify the 3D reconstruction, and compared the method of the present invention with the Lambert model method and the Ward mixture model method.

[0037] [1] Wang Aixue, Zhao Jianhu, Shang Xiaodong et al. Inversion of seafloor 3D microtopography from side-scan sonar images controlled by single-beam bathymetry[J]. Bulletin of Harbin Engineering University, 2017,38(05):739-745. [2] Wang Guohuan, Song Yugui, Ma Fengjun et al. Research progress on shape-from-shading (SFS) technology for non-Lambertian surfaces [J]. Tool Technology, 2012, 46(08): 3-7.

[0038] Figures 3 to 6 show the comparison results of different 3D reconstruction methods for four sets of side-scan sonar images. As shown in Figures 3(a) and 4(a), the two side-scan sonar images of the underwater terrain show a dispersed terrain. Figures 5(a) and 6(a) show two side-scan sonar images of an underwater wind power generation pile foundation, showing a concentrated area of ​​the pile foundation and its surrounding targets.

[0039] As can be seen from Figures 3(b), 4(b), 5(b), and 6(b), when inversion is based on a single diffuse reflectance model, the terrain features are blurred and there are many abnormal protrusions caused by noise, making it impossible to accurately display the relative relief of the original terrain. This is mainly because the general SFS method performs inversion using a single Lambertian diffuse reflectance model, which does not match the actual seafloor surface. As a result, the corresponding features in the 2D side scan image are blurred in the inversion terrain and new noise is introduced, resulting in abnormal protrusions in the inversion terrain and an inaccurate reflection of the actual changes in the seafloor surface relief.

[0040] As can be seen from Figures 3(c), 4(c), 5(c), and 6(c), the terrain features reconstructed using the Ward mixed reflection model are blurred, the topographic relief in local areas is clearly abnormal, and there are several abnormal protrusions on the reconstructed seafloor surface. This is because the Ward mixed reflection model tends to reflect from surfaces with low roughness and cannot adapt to complex seafloor surfaces. Furthermore, because the parameters of the Ward mixed reflection model have a significant impact on the inversion results, setting the parameters based on experience will not achieve an ideal 3D reconstruction result, and will cause abnormal topographic relief in the reconstructed 3D topography and significant differences in features from side-scan sonar images.

[0041] Compared with the above two types of 3D reconstruction methods, the reconstruction results of the method of the present invention, shown in Figures 3(d), 4(d), 5(d), and 6(d), show that the inverse transformation topography clearly displays the variations in the height of the seafloor surface and the topographical features, without any abnormal protrusions. This can effectively and intuitively represent the seafloor topography information in side-scan sonar images. This is mainly because the method of the present invention combines the seafloor surface features near the continental shelf, designs an SFS mixed reflection model for side-scan sonar images, and introduces a Blinn-Phong specular reflection model instead of a single diffuse reflection Lambertian model, and calculates an adaptive specular reflection coefficient D in the form of an exponential function. g and D using a fractional formula including the roughness C of the side scan sonar image. g The purpose is to express the basis number of D and use the roughness C as an index, thereby improving the correlation between the side-scan sonar image and the seafloor surface reflection model. In addition, we design an adaptive specular reflectance exponent ε and g is used as a coefficient in the formula for the specular reflection index ε, and ε is expressed by multiplying the inverse of the roughness by a constant λ, and the adaptive diffuse reflection coefficient D r By designing a SFS mixed reflection model with complex seafloor adaptation, and improving the adaptability of the SFS mixed reflection model, the accuracy of 3D reconstruction of side-scan sonar images can be improved. Compared with other 3D reconstruction methods, the present invention uses the inverse transformation elevation control coefficient k iand a depth data set H={H1, H2..., H i …, H o} and calculate the inverse transformation elevation value Z(N tra / 2,i) to inversely convert the altitude restriction coefficient k i and this coefficient k i By regulating the inverse transformation elevation value Z(x,i) as above, the absolute elevation value Φ(Z(x,i)) is obtained, so that the three-dimensional inverse transformation topography features of the present invention are clear, the changes in relief height match the seafloor features, there are no abnormal protrusions, a high degree of matching is achieved, and the seafloor topography surface is accurately reconstructed in three dimensions.

[0042] To further evaluate the merits of the 3D reconstruction method of the present invention and the comparative methods, we introduced indicators such as mean absolute error (MAE), correlation coefficient (CC), and signal-to-noise ratio (SNR). Here, mean absolute error (MAE) is mainly used to evaluate the error between the 3D projection image and the original image, and the smaller the MAE value, the higher the similarity between the two and the more reasonable the 3D reconstruction of the image.

[0043]

number

[0044] The correlation coefficient (CC) is mainly used to evaluate the similarity between the 3D projection image and the original image, and indicates the ability of the 3D reconstruction algorithm to mine terrain information from side-scan sonar images. The larger the CC value, the better. The specific formula is Equation (17):

number

[0045] The signal-to-noise ratio (SNR) is mainly used to evaluate the degree to which a 3D reconstruction algorithm is affected by image noise. The larger the SNR value, the less the 3D reconstruction algorithm is affected by image noise, and the better its performance.

number

[0046] The 3D reconstruction results for the four types of side-scan sonar images are shown in Table 1. The MAE index of the SFS mixed reflection model 3D reconstruction method described in this specification is reduced by 25.302% and 39.053%, respectively, compared with the Lambert and Ward mixed reflection models. The CC index is improved by 9.590% and 49.649%, respectively, and the SNR index is improved by 23.004% and 31.456%, respectively. This is mainly due to the SFS mixed reflection model of side-scan sonar images designed in this invention. This model introduces the Blinn-Phong specular reflection model to represent the specular reflection portion of the seafloor, forming an SFS mixed reflection model with complex seafloor surface adaptation, which enhances the matching between the reflection model and the seafloor surface. The present invention uses the adaptive specular reflection coefficient D in the form of an exponential function. g and D using a fractional formula including the roughness C of the side scan sonar image. g The base number of the surface is expressed as the roughness C as an index, and the adaptive diffuse reflection coefficient D r and adaptive specular reflectance exponent ε are designed to improve the correlation between the side scan sonar image and the seafloor surface reflectance model, further reducing the MAE index of the method of the present invention and improving the accuracy of 3D reconstruction of the side scan sonar image. i and this coefficient k iBy restricting the inverse transformation elevation value Z(x,i) to , the absolute elevation value Φ(Z(x,i)) is obtained, which reduces the dependency of the 3D reconstruction of side-scan sonar images on external side-scan data, improves the effect of the mixed reflection model in the 3D reconstruction, makes the 3D reconstructed terrain more reasonable, and avoids the introduction of new noise in the 3D reconstruction process, thereby improving the SNR index of the method of the present invention and improving the efficiency and accuracy of 3D reconstruction of side-scan sonar images. From the above, comparing the 3D reconstruction results of four types of side-scan sonar images further verifies that the method of the present invention can effectively achieve a high degree of matching and accurate 3D reconstruction of the seafloor topography.

[0047] [Table 1]

[0048] The above examples are only for illustrating the technical solutions of the present invention and are not limited thereto. Although the present invention has been described in detail with reference to the above examples, those skilled in the art may still make modifications or equivalent substitutions to the specific embodiments of the present invention, and any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention shall fall within the protection scope of the claims of the present invention.

Claims

1. Step (1) of reading a side scan sonar image of a target for three-dimensional reconstruction; Adaptive specular reflection coefficient D ｇ and adaptive diffuse reflection coefficient D ｒ (2) designing an SFS mixed reflection model for side-scan sonar images, including designing an adaptive specular reflectance exponent ε and designing an adaptive seabed surface mixed reflection model formula; The seabed surface elevation value is inversely converted, and the inverse conversion elevation control coefficient k ｉ (3) determining a three-dimensional reconstructed surface elevation, which includes designing and (4) outputting the three-dimensional reconstruction result of the side scan sonar image.

2. The adaptive specular reflection coefficient D ｇ Specifically, designing the following includes steps 1) and 2): 1) Calculate the surface roughness C of the side scan sonar image; 2) Adaptive specular reflection coefficient D ｇ 2. The method for three-dimensional reconstruction of a side-scan sonar image according to claim 1, wherein the following equation (1) is specifically adopted: [0000]

3. The adaptive diffuse reflection coefficient D ｒ 2. The method for reconstructing three-dimensional side-scan sonar images according to claim 1, wherein the design specifically employs equation (2). [Equation 30]

4. The method for three-dimensional reconstruction of side-scan sonar images according to claim 1, characterized in that the designing of the adaptive specular reflectance index ε in step (2) specifically employs equation (3), where λ is a constant. [Equation 31]

5. The step (2) of designing the seabed surface adaptive mixed reflection model equation specifically employs Equation (4): In the formula, R ｈｓ (p, q) is the seabed surface adaptive mixed reflection model function, p and q are the horizontal and vertical gradients of the seabed surface, respectively, φ and τ are the inclination angle and deviation angle of the incident direction of the sonar signal, respectively, ρ is the incident sound wave intensity, and p ｓ , q ｓ 2. The method for three-dimensional reconstruction of a side-scan sonar image according to claim 1, wherein σ and σ are the horizontal and vertical gradients of the energy source position, respectively. [Equation 32]

6. The design of the seabed surface adaptive mixed reflection model equation includes: Specifically, the parameters φ and τ are determined by steps a) to c), a) The direction of the y-axis of the coordinate system at a point on the seabed surface and the direction of the side scan sonar track line are set to be the same. At this time, the incident angle τ of the side scan sonar signal is 0. b) Based on the operating mechanism of the side scan sonar, the tilt angle φ(x, i) corresponding to the x-th pixel in the i-th frame of the side scan sonar image is given by Equation (5): [Equation 33] where arctan(·) is the arctangent function and N ｔｒａ is the horizontal pixel distance of the side scan sonar image, and W ｉ is the seafloor point position in the i-th frame of the side scan sonar image, c) determining the tilt angles of all pixel points in each frame of the side scan sonar image, and further obtaining the corresponding tilt angles of pixel points across the entire side scan sonar image.

7. Inversely converting the seabed surface elevation value in step (3) specifically includes steps (A) to (E), (A) Approximate the seabed surface slopes p and q using the backward finite difference method and discretize them as in equations (6) and (7). [Equation 34] where Z(x, y) is the elevation of the seabed surface at the (x, y) position, p and q are the horizontal and vertical gradients of the seabed surface, respectively. (B) The equation for seabed surface irradiance in a side-scan sonar image is Equation (8), [Equation 35] where R(·) is the reflection model function, (C) Taylor expansion of a pixel point (x, y) on a side scan sonar image gives Equation (9), [Equation 36] (D) Z(x,y)=Z ｎ (x, y), the seabed surface elevation value at the nth iteration is given by equations (10) and (11), [Equation 37] During the ceremony, Z ｎ (x, y), Z ｎ－１ (x, y) are the nth and n-1th iteration results of the seafloor surface elevation, respectively; (E) Set the iteration termination condition μ of the seabed surface elevation, i.e., if Equation (12) is satisfied, the inverse transformation of the seabed surface elevation value is completed; Z ｎ (x,y)-Z ｎ－１ (x,y)≦μ (12) 2. The method for three-dimensional reconstruction of side-scan sonar images according to claim 1, wherein μ is an iteration termination condition for the seabed surface elevation.

8. The inverse transformation altitude restriction coefficient k ｉ Specifically, the designing step (I) and the step (II) are adopted, (I) The seabed point position of the i-th frame of the side scan sonar image is W ｉ If you set it as follows, the current depth H ｉ is Equation (13), [Number 38] In the formula, U is the single-side side scan width of the side scan sonar, and N ｔｒａ is the horizontal pixel distance in the side-scan image, (II) The inverse transformation altitude control coefficient k of the i-th frame of the side scan sonar image below ｉ Designed and [0.39] 2. The method for three-dimensional reconstruction of a side scan sonar image according to claim 1, wherein is an inverse transformation elevation value of the track line position in the i-th frame of the side scan sonar image.

9. The step (3) of determining the three-dimensional reconstructed surface elevation specifically includes steps (a) to (e), (a) Inversely transforming the seafloor surface elevation values ​​of the side scan sonar image; (b) When all depth data on the track line is obtained, the depth set H = {H １ , H ２ …, H ｉ …, H ｏ }, (c) Inverse transformation altitude restriction coefficient k of the i-th frame of the side scan sonar image ｉ Calculate (d) The equation for the inverse transformed elevation value Φ(Z(x,i)) corresponding to the i-th frame of the regulated side scan sonar image is Equation (15): [Equation 40] where Z(x,i) is the inverse transformed elevation value corresponding to the i-th frame of the side scan sonar image; (e) after completing the elevation value regulation for all frames of the side scan sonar image of the i-th frame, elevation value regulation for the side scan sonar images of other frames is sequentially performed, and further, regulation of the seabed surface elevation scale of the side scan sonar image is completed. This is a three-dimensional reconstruction method of a side scan sonar image as described in claim 1.

Images