3D Sonar Image Reconstruction Based on Multilayered Mesh Search and Triangular Connection

3D sonar image reconstruction is the basis of sonar detection. But the distribution of sonar data is stereoscopic and multilayered, and the background echo noise is much. The traditional Mesh search will produce a large number of invalid triangular facets and lots of holes will be formed. The paper presents a new sonar image reconstruction method. In order to achieve the surface reconstruction of sonar target efficiently, the method firstly performs 2×2 grid block search for nonempty sonar data points using the adaptive threshold, and then searches for the 3×3 grid blocks with the empty point as the center to reduce the sonar image holes. Then the system performs 3×2 horizontal grid blocks and 2×3 vertical grid blocks search for sonar arrays to further improve the connectivity relation by finding the semi-diagonal connection. Finally, a triangular connection and reconstruction are performed on the basis of the obtained sonar data point connectivity. Experimental results show that the proposed method can correctly reflects the underwater terrain.