基于前视声呐的水下目标检测算法研究
2022,30(11):17-24
摘要：水声目标检测就是从水声取目标信息并进行识别,而有效的水声标检测在现代化的海洋开发中有着十分重要的作用；首先,介绍了水声目标检测所采用的设备,以及该设备的工作方式,并搭建试验场景进行水声回波的采集；其次,对水声标检测算法进行了研究,比对不同的滤波算法可知,中值滤波方法的去噪效果更佳；进行了灰度直方图分析,对目标的灰度范围进行了增强处理；对图像分割算法进行了研究,设计了一种自适应迭代分割算法,对比其他方法有着更好的处理效果；对分割后的图像进行连通域查找,目标筛选,从而识别出待检测目标；试验结果验证了该水声目标检测算法在对水下目标的检测识别上的有效性。
关键词：声呐图像;水声目标;水下目标检测;图像分割
Research on Underwater Target Detection Algorithm Based on Forward Looking Sonar
Abstract：Underwater acoustic target detection is to take target information from underwater acoustic and identify it, and effective underwater acoustic beacon detection plays a very important role in modern ocean development. Firstly, the equipment used in underwater acoustic target detection and its working mode are introduced, and a test scene is set up to collect underwater acoustic echo. Secondly, the detection algorithm of hydroacoustic beacon is studied. Compared with different filtering algorithms, the median filtering method has better denoising effect. The gray histogram is analyzed, and the gray range of the target is enhanced. The image segmentation algorithm is studied, and an adaptive iterative segmentation algorithm is designed, which has better processing effect than other methods. Search the connected domain of the segmented image and screen the target, so as to identify the target to be detected; The experimental results verify the effectiveness of the underwater acoustic target detection algorithm in the detection and recognition of underwater targets.
Key words：Sonar image;Underwater acoustic target;Underwater target detection;Image segmentation
收稿日期：2022-03-30
基金项目：本文部分受国家自然科学基金(61903304),和111引智项目(B18041)支持。
