A method for fusing airborne LiDAR bathymetry and multi-beam sonar bathymetry point cloud

Abstract

The present invention discloses a method for fusing airborne LiDAR bathymetry and multi-beam sonar bathymetry point clouds, belonging to the field of surveying and mapping technology. The present invention first divides the original point cloud into overlapping areas and non-overlapping areas using a virtual grid, and generates simulated grid nodes with certain noise resistance, and then constructs a seamless integration model based on non-overlapping blocks, locally aligns the source measurement points in each non-overlapping block with the corresponding simulated grid nodes, and uses the TPS interpolation algorithm to smooth the boundary points of each non-overlapping block. Through this method, the present invention can reduce terrain distortion to a certain extent, eliminate data gaps in the seam area of ​​airborne LiDAR bathymetry and multi-beam sonar bathymetry point cloud data, and provide a data basis for applications such as three-dimensional reconstruction of point cloud data.

Description

Technical Field

[0001] The present invention belongs to the field of surveying and mapping technology, and in particular relates to a method for fusing airborne LiDAR depth measurement with multi-beam sonar depth measurement point clouds. Background Art

[0002] Airborne LiDAR Bathymetry (ALB) and Multibeam Echo Sounder (MBES) are two active remote sensing technologies that have developed rapidly in recent decades. ALB and MBES systems integrate inertial navigation systems (INS) and global navigation satellite systems (GNSS), so they can directly obtain discrete, large-scale, high-resolution three-dimensional point clouds of targets. However, due to the limitations of the survey range of ALB and MBES, it is difficult for a single sensor to obtain complete information about the target. Specifically, due to the limitations of the range of ship activities, MBES cannot obtain information on land, shallow water, and reef areas; on the other hand, ALB can obtain surface and shallow water information, but cannot obtain deep water and turbid water information. Since the two technologies are complementary in terms of their survey range, it is necessary to integrate ALB and MBES data to provide complete above-water and underwater integrated information of the target scene. Since non-homologous point cloud data may have certain differences in the digital description of the same object, and are affected by factors such as registration errors and sensor system errors, there are often certain gaps at the seams of the two point cloud data, which are difficult to be directly used for three-dimensional reconstruction and other tasks. Summary of the invention

[0003] In view of the above technical problems existing in the prior art, the present invention proposes a method for fusing airborne LiDAR bathymetry and multi-beam sonar bathymetry point cloud, which has a reasonable design, overcomes the shortcomings of the prior art and has good effects.

[0004] In order to achieve the above object, the present invention adopts the following technical solution:

[0005] A method for fusing airborne LiDAR bathymetry and multi-beam sonar bathymetry point clouds comprises the following steps:

[0006] Step 1: Use the ICP algorithm to initially align the airborne LiDAR bathymetry and multi-beam sonar bathymetry point clouds;

[0007] Step 2: Construct a virtual grid covering the entire study area according to the preset virtual grid cell size parameter cs, so as to identify the overlapping and non-overlapping areas of the airborne LiDAR bathymetry and multi-beam sonar bathymetry point clouds;

[0008] Step 3: Build a seamless integration model based on non-overlapping blocks to eliminate the data gap at the junction of airborne LiDAR bathymetry and multi-beam sonar bathymetry point clouds.

[0009] Preferably, in step 2, the steps of identifying overlapping and non-overlapping areas are as follows:

[0010] Step 2.1: Use different symbols to mark the airborne LiDAR bathymetry and multibeam sonar bathymetry point clouds to distinguish between the two;

[0011] Step 2.2: Combine the airborne LiDAR bathymetry and multi-beam sonar bathymetry point clouds, and count the maximum value of the plane coordinates of all measurement points, i.e., x max and max With the minimum value x min and min , in order to construct the minimum enclosing rectangle of the merged point cloud on the horizontal plane; then, a virtual grid covering the entire study area is constructed according to the preset virtual grid unit size parameter cs; according to the positional relationship between each measurement point of the merged point cloud and the center of the virtual grid unit, the virtual grid unit index number corresponding to each measurement point is determined; specifically, the number of rows of the virtual grid N row With the number of columns N column Calculated by the following formula:

[0012]

[0013] Among them, Floor(h) means rounding down, that is, obtaining an integer not greater than h; Ceil(h) means rounding up, that is, obtaining an integer not less than h; x max and max is the maximum value of the plane coordinate; x min and min is the minimum value of the plane coordinate;

[0014] Then, the plane position (x) of the center of each virtual grid cell is calculated using the following formula: center ,y center ):

[0015]

[0016] Among them, i x =0,1,...,N column ,i y =0,1,…,N row ;

[0017] The index number (x idx ,y idx ), so that each measurement point corresponds to a virtual grid unit, thereby dividing the merged point cloud data into the corresponding virtual grid units:

[0018]

[0019] Among them, Round(h) means rounding h;

[0020] Step 2.3: For a non-empty virtual grid cell, if it contains both airborne LiDAR bathymetry and multi-beam sonar bathymetry point clouds, the virtual grid cell is considered to be an overlapping area, and the measurement points therein are marked as overlapping ALB measurement points and overlapping MBES measurement points; if only airborne LiDAR bathymetry point clouds exist in the virtual grid cell, the virtual grid cell is considered to be a non-overlapping area, and the measurement points therein are marked as non-overlapping ALB measurement points; traverse all grid cells to realize the identification of overlapping and non-overlapping areas.

[0021] Preferably, in step 3, the steps of constructing the seamless integration model of non-overlapping blocks are as follows:

[0022] Step 3.1: Generate simulated grid nodes using the overlapping MBES measurement points in the virtual grid cells, construct a temporary DTM through the generated simulated grid nodes, and calculate the normal vector at the position of each simulated grid node;

[0023] Step 3.2: Use the point cloud segmentation algorithm based on Euclidean distance to segment the non-overlapping ALB measurement points into non-intersecting segmentation units;

[0024] Step 3.3: For a segmentation unit, if the number of measurement points in it is greater than a certain threshold, the α-shape algorithm is used to extract its contour points, and then a buffer of a certain width is constructed outside the segmentation unit along the segmentation unit contour points, and the simulated grid nodes and overlapping ALB measurement points in the buffer are extracted; otherwise, the next segmentation unit is processed;

[0025] Step 3.4: According to the obtained virtual grid cell index number, the overlapping ALB measurement points are matched with the simulated grid nodes one by one;

[0026] Step 3.5: Taking the simulated grid nodes as reference, the transformation parameters between the overlapping ALB measurement points and the simulated grid nodes are calculated using the point-to-surface transformation model and the stepwise approach method, and then the transformation parameters are applied to the segmentation units and the overlapping ALB measurement points, and then the two are merged;

[0027] Step 3.6: Repeat steps 3.3 to 3.5 until all segmentation units are processed;

[0028] Step 3.7: Use the virtual grid again to extract non-overlapping ALB measurement points in the merged point cloud, and extract their contour points using the α-shape algorithm;

[0029] Step 3.8: Calculate the distance between the contour point and the non-overlapping ALB measurement point, mark the non-overlapping ALB measurement points less than a certain distance threshold, including the contour point itself, as points to be smoothed, and then remove them from the non-overlapping ALB measurement point set;

[0030] Step 3.9: For a point to be smoothed, search for its nearest k simulated grid nodes and k non-overlapping ALB measurement points, and generate a smooth surface in combination with the TPS interpolation algorithm. Move the point to be smoothed along the vertical direction to the resulting smooth surface; traverse all points to be smoothed to complete the gap elimination.

[0031] Beneficial technical effects brought by the present invention:

[0032] The present invention proposes a method for fusing airborne LiDAR bathymetry and multi-beam sonar bathymetry point clouds. Compared with the prior art, the present invention uses a virtual grid to identify overlapping and non-overlapping areas; constructs a seamless integration model based on non-overlapping blocks; and can reduce terrain distortion to a certain extent and eliminate data gaps in non-homologous data seam areas. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 A flow chart of a method for fusing airborne LiDAR bathymetry and multi-beam sonar bathymetry point cloud according to the present invention;

[0034] Figure 2 A schematic diagram of identifying overlapping and non-overlapping areas in the present invention;

[0035] in, Figure 2 (a) Schematic diagram of virtual grid construction; Figure 2 (b) is the result of overlapping and non-overlapping area recognition;

[0036] Figure 3 It is a schematic diagram of a seamless integration model based on non-overlapping blocks in the present invention;

[0037] in, Figure 3 (a) Schematic diagram of minimizing the distance from a point to the temporary DTM surface; Figure 3 (b) is the local alignment result diagram; Figure 3 (c) Schematic diagram of the smooth surface of TPS; Figure 3 (d) is the seamless integration result diagram. DETAILED DESCRIPTION

[0038] The present invention is further described in detail below with reference to the accompanying drawings and specific embodiments:

[0039] The present invention provides a method for fusing airborne LiDAR bathymetry and multi-beam sonar bathymetry point cloud, the process of which is as follows: Figure 1 As shown, the following steps are included:

[0040] Step 1: Use the ICP algorithm to initially align the airborne LiDAR bathymetry and multi-beam sonar bathymetry point clouds;

[0041] Step 2: Construct a virtual grid covering the entire study area according to the preset virtual grid cell size parameter cs, so as to identify the overlapping and non-overlapping areas of the airborne LiDAR bathymetry and multi-beam sonar bathymetry point clouds;

[0042] In a further embodiment, step 2 specifically includes the following steps:

[0043] Step 2.1: Use different symbols to mark the airborne LiDAR bathymetry and multibeam sonar bathymetry point clouds to distinguish between the two;

[0044] Step 2.2: Merge the airborne LiDAR bathymetry and multi-beam sonar bathymetry point clouds, and count the maximum value of the plane coordinates of all measurement points (x max and max ) and the minimum value (x min and min ) to construct the minimum enclosing rectangle of the merged point cloud on the horizontal plane; then a virtual grid covering the entire study area is constructed according to the preset virtual grid unit size parameter cs, such as Figure 2 As shown in (a); according to the positional relationship between each measurement point of the merged point cloud and the center of the virtual grid unit, the virtual grid unit index number corresponding to each measurement point is determined.

[0045] Specifically, the number of rows N of the virtual grid row With the number of columns N column It can be calculated by the following formula:

[0046]

[0047] Among them, Floor(h) means rounding down, that is, obtaining an integer not greater than h; Ceil(h) means rounding up, that is, obtaining an integer not less than h. Then, the plane position (x) of the center of each virtual grid unit is calculated using the following formula: center ,y center ):

[0048]

[0049] Among them, i x =0,1,...,N column ,i y =0,1,…,N row The index number (x idx ,y idx), so that each measurement point corresponds to a virtual grid unit, thereby dividing the merged point cloud data into the corresponding virtual grid units:

[0050]

[0051] Among them, Round(h) means rounding h.

[0052] Step 2.3: For a non-empty virtual grid cell, if it contains both airborne LiDAR bathymetry and multi-beam sonar bathymetry point clouds, the virtual grid cell is considered to be an overlapping area, and the measurement points therein are marked as overlapping ALB measurement points and overlapping MBES measurement points; if the virtual grid cell only contains airborne LiDAR bathymetry point clouds, the virtual grid cell is considered to be a non-overlapping area, and the measurement points therein are marked as non-overlapping ALB measurement points. Traverse all grid cells to identify overlapping and non-overlapping areas, such as Figure 2 (b) as shown.

[0053] Step 3: Build a seamless integration model based on non-overlapping blocks to eliminate the data gap at the junction of airborne LiDAR bathymetry and multi-beam sonar bathymetry point clouds.

[0054] In a further embodiment, step 3 specifically includes the following steps:

[0055] Step 3.1: Generate simulated grid nodes using the overlapping MBES measurement points in the virtual grid cells, construct a temporary DTM through the generated simulated grid nodes, and calculate the normal vector at the position of each simulated grid node;

[0056] Step 3.2: Use the point cloud segmentation algorithm based on Euclidean distance to segment the non-overlapping ALB measurement points into non-intersecting segmentation units;

[0057] Step 3.3: For a segmentation unit, if the number of measurement points in it is greater than a certain threshold (such as 30), the α-shape algorithm is used to extract its contour points, and then a buffer of a certain width is constructed outside the segmentation unit contour points, and the simulated grid nodes and overlapping ALB measurement points in the buffer are extracted; otherwise, the next segmentation unit is processed;

[0058] Step 3.4: According to the obtained virtual grid cell index number, the overlapping ALB measurement points are matched with the simulated grid nodes one by one;

[0059] Step 3.5: Taking the simulated grid nodes as reference, the transformation parameters between the overlapping ALB measurement points and the simulated grid nodes are calculated using the point-to-surface transformation model and the stepwise approach method, and then the transformation parameters are applied to the segmentation units and the overlapping ALB measurement points, and the two are merged, as shown in Figure 3As shown in (a) and (b);

[0060] Step 3.6: Repeat steps 3.3 to 3.5 until all segmentation units are processed;

[0061] Step 3.7: Use the virtual grid again to extract non-overlapping ALB measurement points in the merged point cloud, and extract their contour points using the α-shape algorithm;

[0062] Step 3.8: Calculate the distance between the contour point and the non-overlapping ALB measurement point, mark the non-overlapping ALB measurement points (including the contour point itself) that are less than a certain distance threshold as points to be smoothed, and then remove them from the non-overlapping ALB measurement point set;

[0063] Step 3.9: For a point to be smoothed, search for its nearest k simulated grid nodes and k non-overlapping ALB measurement points, and generate a smooth surface by combining the TPS interpolation algorithm. Move the point to be smoothed along the vertical direction to the resulting smooth surface, as shown in Figure 3 As shown in (c) and (d), all the points to be smoothed are traversed to complete the gap elimination.

[0064] In specific implementation, the invented airborne LiDAR bathymetry and multi-beam sonar bathymetry point cloud fusion method can effectively reduce terrain distortion and eliminate data gaps in the seam area of ​​airborne LiDAR bathymetry and multi-beam sonar bathymetry point cloud data.

[0065] In summary, the present invention provides a method for fusing airborne LiDAR bathymetry and multi-beam sonar bathymetry point clouds, the method comprising: using a virtual grid to divide the original point cloud into overlapping areas and non-overlapping areas, and generating simulated grid nodes with certain noise resistance, and then constructing a seamless integration model based on non-overlapping blocks, locally aligning the source measurement points in each non-overlapping block with the corresponding simulated grid nodes, and using the TPS interpolation algorithm to smooth the boundary points of each non-overlapping block. The present invention effectively solves the data gap problem in the seam area of ​​the airborne LiDAR bathymetry and multi-beam sonar bathymetry point cloud data, and realizes the high-precision fusion of airborne LiDAR bathymetry and multi-beam sonar bathymetry point clouds.

[0066] Of course, the above description is not a limitation of the present invention, and the present invention is not limited to the above examples. Changes, modifications, additions or substitutions made by technicians in this technical field within the essential scope of the present invention should also fall within the protection scope of the present invention.

Claims

1. A method for fusing airborne LiDAR bathymetry and multi-beam sonar bathymetry point clouds, characterized by: The following steps are involved: Step 1: Use the ICP algorithm to initially align the airborne LiDAR bathymetry and multi-beam sonar bathymetry point clouds; Step 2: Construct a virtual grid covering the entire study area according to the preset virtual grid cell size parameter cs, so as to identify the overlapping and non-overlapping areas of the airborne LiDAR bathymetry and multi-beam sonar bathymetry point clouds; Step 3: Build a seamless integration model based on non-overlapping blocks to eliminate the data gap at the junction of the airborne LiDAR bathymetry and multi-beam sonar bathymetry point clouds; In step 2, the steps of identifying overlapping and non-overlapping areas are as follows: Step 2.1: Use different symbols to mark the airborne LiDAR bathymetry and multibeam sonar bathymetry point clouds to distinguish between the two; Step 2.2: Combine the airborne LiDAR bathymetry and multi-beam sonar bathymetry point clouds, and count the maximum value of the plane coordinates of all measurement points, i.e., x max and max With the minimum value x min and min , in order to construct the minimum enclosing rectangle of the merged point cloud on the horizontal plane; then, a virtual grid covering the entire study area is constructed according to the preset virtual grid unit size parameter cs; according to the positional relationship between each measurement point of the merged point cloud and the center of the virtual grid unit, the virtual grid unit index number corresponding to each measurement point is determined; specifically, the number of rows of the virtual grid N row With the number of columns N column Calculated by the following formula: Among them, Floor(h) means rounding down, that is, obtaining an integer not greater than h; Ceil(h) means rounding up, that is, obtaining an integer not less than h; x max and max is the maximum value of the plane coordinate; x min and min is the minimum value of the plane coordinate; Then, the plane position (x) of the center of each virtual grid cell is calculated using the following formula: center ,y center ): Among them, i x = 0, 1, ..., N column , i y = 0, 1, …, N row ; The index number (x idx ,y idx ), so that each measurement point corresponds to a virtual grid unit, thereby dividing the merged point cloud data into the corresponding virtual grid units: Among them, Round(h) means rounding h; Step 2.3: For a non-empty virtual grid cell, if it contains both airborne LiDAR bathymetry and multi-beam sonar bathymetry point clouds, the virtual grid cell is considered to be an overlapping area, and the measurement points therein are marked as overlapping ALB measurement points and overlapping MBES measurement points; if only airborne LiDAR bathymetry point clouds exist in the virtual grid cell, the virtual grid cell is considered to be a non-overlapping area, and the measurement points therein are marked as non-overlapping ALB measurement points; traverse all grid cells to realize the identification of overlapping and non-overlapping areas.

2. The method for fusing airborne LiDAR bathymetry and multi-beam sonar bathymetry point cloud according to claim 1, characterized in that: In step 3, the steps for building a seamless integration model for non-overlapping blocks are as follows: Step 3.1: Generate simulated grid nodes using the overlapping MBES measurement points in the virtual grid cells, construct a temporary DTM through the generated simulated grid nodes, and calculate the normal vector at the position of each simulated grid node; Step 3.2: Use the point cloud segmentation algorithm based on Euclidean distance to segment the non-overlapping ALB measurement points into non-intersecting segmentation units; Step 3.3: For a segmentation unit, if the number of measurement points in it is greater than a certain threshold, the α-shape algorithm is used to extract its contour points, and then a buffer of a certain width is constructed outside the segmentation unit along the segmentation unit contour points, and the simulated grid nodes and overlapping ALB measurement points in the buffer are extracted; otherwise, the next segmentation unit is processed; Step 3.4: According to the obtained virtual grid cell index number, the overlapping ALB measurement points are matched with the simulated grid nodes one by one; Step 3.5: Taking the simulated grid nodes as reference, the transformation parameters between the overlapping ALB measurement points and the simulated grid nodes are calculated using the point-to-surface transformation model and the stepwise approach method, and then the transformation parameters are applied to the segmentation units and the overlapping ALB measurement points, and then the two are merged; Step 3.6: Repeat steps 3.3 to 3.5 until all segmentation units are processed; Step 3.7: Use the virtual grid again to extract non-overlapping ALB measurement points in the merged point cloud, and extract their contour points using the α-shape algorithm; Step 3.8: Calculate the distance between the contour point and the non-overlapping ALB measurement point, mark the non-overlapping ALB measurement points less than a certain distance threshold, including the contour point itself, as points to be smoothed, and then remove them from the non-overlapping ALB measurement point set; Step 3.9: For a point to be smoothed, search for its nearest k simulated grid nodes and k non-overlapping ALB measurement points, and generate a smooth surface by combining the TPS interpolation algorithm, and move the point to be smoothed along the vertical direction to the resulting smooth surface; Traverse all points to be smoothed and eliminate gaps.

Images