| With the broad application of underwater acoustic exploration in ocean resource development, sonar imaging technology has become an important method of underwater detection. However, now the general way of3D sonar imaging is using PC platform, thus results in complex scales and low integration. Although there are some forward looking sonar imaging systems, the size and power consumption of them are large, and the visualization is poor, so they are not suitable to be used underwater for frogman. Therefore, studying hand-held3D sonar imaging system has high value in engineering.As traditional3D sonar imaging systems have the problems of low integration and poor real-time, this thesis designed and implemented software for imaging system of real-time hand-held3D sonar, and the module of signal transmission and acquisition is integrated with the image processing module, making the system suitable for frogman. This system uses DM8127to realize sonar image processing, and the image processing has three main steps:data analysis, triangular mesh construction and image rendering. For this system runs in a modular way, there designs a communication mechanism between modules to achieve message interaction. Meanwhile, in consideration of the limited memory capacity and bandwidth of embedded systems, there are several means designed to achieve high-precision and real-time3D modeling, such as pre-generation of coordinate conversion array, mesh construction based on adaptive threshold and multi-thread parallel processing, and using OpenGL ES2.0to achieve the rendering and display of3D sonar images. Experimental results showed that this system can clearly show the underwater target topography, and the image processing speed can reach up to25frames per second. This system has met the requirements of good real-time and high integration. |