With the continuous research in the underwater scene exploration, 3D imaging sonar could get an intuitive target image compared with 2D imaging sonar. The sonar device emits sound waves and acquires position and image based on the echo signals. However, the huge amount of data per frame and the environmental interference bring great difficulty to the recognition and visualization of the target.
In order to obtain the true topological structure of the target, researchers from the Institute of Acoustics of the Chinese Academy of Sciences (IACAS) proposed an improved forward-looking sonar 3D Visualization Scheme which significantly improved the visual effect of underwater target.
The study has been published online in IEEE Access.
The study carried out echo simulation and 3D imaging of targets in a shallow water environment with reverberation. Afterwards, aiming at the noise and abnormal points caused by environmental interference, researchers proposed a global iterative threshold segmentation method based on Local Outlier Factor (LOF). The idea of data mining was applied to the three-dimensional imaging sonar signal processing for the first time, removing interference information while maintaining the target characteristics. Finally, they proposed a Power Crust reconstruction algorithm based on Random Sample Consensus (RANSC) to improve the visual effect of target reconstruction.
For the echo data acquired by the three-dimensional sonar, it was difficult to directly identify the target. Through this improved visualization scheme, the research greatly improved the visual effect of the target and reduced the amount of data. In the future, this research could draw obstacles for underwater submersibles and scenes of the marine environment, improve the accuracy of implementation, and bring more realistic and intuitive feelings.
Figure 1. Forward-looking sonar 3D visualization scheme flow chart. (Image by IOA)
Figure 2. Raw data (left) and target topology (right). (Image by IOA)
The research was supported by Youth Innovation Promotion Association of the Chinese Academy of Sciences.
Teng Zeng, Chunhua Zhang, Yong Huang, Peng Wang. An Improved Forward-Looking Sonar 3D Visualization Scheme of Underwater Objects. IEEE Access, 2019, 7: 147983-147990. DOI: 10.1109/ACCESS.2019.2946343
Institute of Acoustics, Chinese Academy of Sciences, 100190 Beijing, China