CN-121995526-A - Three-dimensional comprehensive base map construction method for underwater small target detection

CN121995526ACN 121995526 ACN121995526 ACN 121995526ACN-121995526-A

Abstract

The invention relates to a three-dimensional comprehensive base map construction method for underwater small target detection, which comprises the following steps of formulating a hierarchical grid segmentation coding rule according to reference submarine topography data resolution and detection sonar data resolution, and constructing a submarine three-dimensional base map grid model capable of realizing no-precision loss and variable granularity updating according to the coding rule, wherein vertical offset of grid vertexes is used for bearing and reflecting elevation data acquired by depth sonar, and attribute numerical values attached to the grid vertexes are used for bearing scanned region identification information and acoustic detection intensity data acquired by side scan sonar. The invention can construct three-dimensional terrain by using open source terrain data in a lossless manner, comprehensively express various sonar detection data on a unified base map in real time, and form capability complementation among underwater small target detection equipment.

Inventors

JIN CHAOCHAO
ZHANG WEI
LI YANAN
JIN XIAOFAN

Assignees

中国船舶集团有限公司第七一五研究所

Dates

Publication Date: 20260508
Application Date: 20260202

Claims (8)

1. A three-dimensional comprehensive base map construction method for underwater small target detection is characterized by comprising the following steps of, According to the reference submarine topography data resolution and the detection sonar data resolution, a hierarchical grid segmentation coding rule is formulated; According to the coding rule, constructing a submarine three-dimensional base map grid model capable of realizing no precision loss and variable granularity updating, wherein the vertical offset of grid vertexes is used for bearing and reflecting elevation data acquired by depth sounding sonar, and attribute values attached to the grid vertexes are used for bearing scanned region identification information and acoustic detection intensity data acquired by side-scan sonar; and acquiring elevation offset data and incidental attribute values of the grid vertexes in real time through a shader, carrying out fusion rendering on the reference terrain, the sounding fine terrain and the acoustic detection intensity image, and generating and updating a three-dimensional comprehensive base map.
2. The three-dimensional comprehensive base map construction method for underwater small target detection according to claim 1, wherein the establishment of the hierarchical grid segmentation coding rule is specifically as follows: Respectively constructing grid subdivision arrays in the longitudinal direction and the latitudinal direction according to the preset number of the single grid units; ensuring that the grid spacing of the level at which the initial terrain grid is located is the same as the resolution of the reference seafloor terrain data.
3. The method for constructing the three-dimensional comprehensive bottom map for underwater small target detection according to claim 1, wherein the method for constructing the three-dimensional bottom map grid model of the seabed is specifically as follows: instantiating an initial grid according to a hierarchy corresponding to the reference seabed topography data for a working area; the three-way offset and scaling parameters of each mesh vertex are matched with the data points of the reference seafloor terrain data.
4. The method for constructing a three-dimensional comprehensive base map for underwater small target detection according to claim 1, wherein the depth sounding sonar elevation data reflected by grid vertex offset is specifically: determining a corresponding grid subdivision level according to the depth sounding sonar data resolution; And positioning to the corresponding grid vertex with the unique code according to the coordinates of the depth finding sonar point cloud data, and adjusting the vertical offset of the vertex.
5. The method for constructing a three-dimensional comprehensive base map for underwater small target detection according to claim 1, wherein the processing of acoustic detection intensity data is specifically as follows: Based on echo intensity data of the side scan sonar, combining navigation and attitude data of the detection platform, and processing and generating associated data of geographic coordinates and echo intensity; and positioning to the corresponding grid vertex with the unique code according to the geographic coordinates in the associated data, and adjusting the appointed attribute value attached to the vertex according to the echo intensity.
6. The three-dimensional comprehensive base map construction method for underwater small object detection according to claim 1, wherein the attributes attached to the grid vertices comprise basic attributes for constructing terrains and extended attribute parameters which can be called by a shader, and the extended attribute parameters comprise R, G, B, A components for color rendering and texture coordinate parameters u and v.
7. The method for constructing the three-dimensional comprehensive base map for underwater small target detection according to claim 1, wherein the step of completing real-time three-dimensional comprehensive base map rendering through the shader comprises the following steps: the shader is used for calling the vertical coordinates of the grid vertexes and the orientation information of the patches to render the submarine topography; The shader invokes the grid vertex attribute parameters associated with the sounding sonar data to render and indicate the scanned area; The shader invokes the mesh vertex attribute parameters associated with the echo intensity data to render the acoustic image.
8. The method for constructing the three-dimensional comprehensive base map for underwater small target detection according to claim 1, wherein the three-dimensional comprehensive base map rendering is characterized in that in the same three-dimensional scene, the reference topographic data, the fine topographic data acquired by depth sounding sonar and the echo intensity image data acquired by side scan sonar are synchronously overlapped and fused for display.

Description

Three-dimensional comprehensive base map construction method for underwater small target detection Technical Field The invention belongs to the technical field of underwater sound detection and application, and particularly relates to a three-dimensional comprehensive base map construction method for underwater small target detection. Background The underwater small target detection is mainly oriented to threat targets such as submerged buoy, frogman, underwater mine, unmanned autonomous vehicle and the like in the sea. The main detection means at present comprise multi-beam sonar, side-scan sonar, synthetic aperture sonar, three-dimensional imaging sonar and the like. The multi-beam sonar, the three-dimensional imaging sonar and other sounding sonars can acquire the relative angle and distance of the arrival position of the beam, the submarine topography point cloud data can be obtained by combining the platform navigation and the gesture correction, the suspended target can be identified by analyzing the abnormal protrusion of the point cloud, but the bottom-attached target cannot be identified, and the beam still only can identify the seabed due to the condition that the echo intensity of the target is weaker than that of the seabed echo. Sonar images of side-scan sonar, synthetic aperture sonar and the like mainly reflect the change relation of echo intensity along with receiving time (inclined distance) after sound waves are transmitted, and targets with higher echo intensity can form bright spots on the sonar images, but as the data do not contain elevation information, submarine topography protrusions cannot be distinguished, and in-water suspended targets cannot be effectively identified. At present, most underwater small-target detection sonar equipment is independently used, effective capability complementation is not formed, for example, the topographic features of multi-beam sonar mapping can identify the image features of side-scan sonar, and the side-scan sonar can make up for the defects that the multi-beam sonar cannot identify the bottom-attached target, the weak-intensity target is lost and the like. In the aspect of data application, a large amount of sonar data generally depends on post-processing to construct data products, and instant detection operation cannot be effectively supported. On the other hand, open source and history detected topographic data are not effectively utilized for detecting and identifying small underwater targets in increment, and the conventional methods for constructing three-dimensional topography through rasterization such as quadtrees and the like often lose the accuracy of the source data. Therefore, there is a need for an underwater three-dimensional comprehensive base map construction method that can nondestructively fuse multi-source data, and support real-time rendering and target recognition. Disclosure of Invention The invention aims to solve the technical problem of providing a three-dimensional comprehensive base map construction method for underwater small target detection, which can construct three-dimensional terrains by using open source terrains without damage, comprehensively express various sonar detection data on a unified base map in real time and form capability complementation among underwater small target detection equipment. The technical proposal of the invention is to provide a three-dimensional comprehensive base map construction method for underwater small target detection, which comprises the following steps, According to the reference submarine topography data resolution and the detection sonar data resolution, a hierarchical grid segmentation coding rule is formulated; According to the coding rule, constructing a submarine three-dimensional base map grid model capable of realizing no precision loss and variable granularity updating, wherein the vertical offset of grid vertexes is used for bearing and reflecting elevation data acquired by depth sounding sonar, and attribute values attached to the grid vertexes are used for bearing scanned region identification information and acoustic detection intensity data acquired by side-scan sonar; and acquiring elevation offset data and incidental attribute values of the grid vertexes in real time through a shader, carrying out fusion rendering on the reference terrain, the sounding fine terrain and the acoustic detection intensity image, and generating and updating a three-dimensional comprehensive base map. Preferably, the formulation of the hierarchical grid partition coding rule specifically includes: Respectively constructing grid subdivision arrays in the longitudinal direction and the latitudinal direction according to the preset number of the single grid units; ensuring that the grid spacing of the level at which the initial terrain grid is located is the same as the resolution of the reference seafloor terrain data. Preferably, the construction of the three-dimensional bottom map