Improved Identification Evaluation of Side-Scan Sonar for Blue Carbon

Sonar can only provide information about the presence of objects at a given location. Therefore, the evaluation accuracy of sonar is lower than that of cameras and is not often used for identification of living organisms. The objective of this research is to improve the evaluation system by using sonar to identify the differences among amamo and sasabamo, which are surveyed as blue carbon, and waterweeds that are not surveyed, including ebimo, respectively. Sonar data were acquired by boat or on foot. The acquired data is Fourier transformed using MATLAB. The difference between the Fourier transformed images at each location is detected. The detected results are compared with the images taken by the underwater drone that was operated together with the survey to improve the accuracy. It is difficult to detect plants using only sonar data showing acoustic reflection intensity and underwater drone footage captured in turbid water with overexposure. Fourier transforms were performed for areas where plants were inhabited and areas where plants were not inhabited. The comparison of the two results revealed findings suggesting that differences can be observed even in areas with similar seabed conditions.