A kind of method and system for three-dimensional modeling of deep water net cage fish population based on sonar
By deploying multibeam sonar on the top and sides of the net cage and combining global and local point cloud data, the problem of detection blind spots in the construction of three-dimensional models of deep-sea aquaculture fish schools was solved, realizing accurate reconstruction and visualization of the three-dimensional model of the fish school and supporting refined management.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- GUANGDONG OCEAN UNIVERSITY
- Filing Date
- 2025-12-11
- Publication Date
- 2026-07-24
AI Technical Summary
Existing sonar detection methods cannot solve the problems of limited detection range and blind spots, resulting in inaccurate construction of three-dimensional models of deep-sea aquaculture fish schools, which makes it impossible to achieve refined and intelligent management.
Multibeam sonar is deployed on the top of the cage to acquire global 3D point cloud data, and multibeam sonar probes are deployed on the sides of key depth layers to acquire local 2D point cloud data. The fusion mechanism of global 3D point cloud and local 2D point cloud is combined to perform 3D reconstruction and visualization.
It achieves completeness and accuracy in the three-dimensional model of fish swarms in deep-sea cage culture, improves the reliable analysis capabilities of fish population, distribution, behavior and health status, and supports accurate assessment and safety management.
Smart Images

Figure CN121661257B_ABST