声学摄像仪图像处理的初步研究及在渔业上的应用  

Preliminary Research on the Image Processing of Acoustic Camera and Its Application in Fishery

作　　者：童剑锋[1] 韩军[2,1] 沈蔚[1] 

机构地区：[1]上海海洋大学海洋科学学院,上海201306 [2]东京大学生产技术研究所

出　　处：《湖南农业科学》2010年第9期149-152,156,共5页Hunan Agricultural Sciences

基　　金：上海市科委科技攻关项目(编号:075905112)

摘　　要：声学摄像仪DIDSON(Dual-frequency IDentification SONar)可以在黑暗、浑浊的水中提供接近光学画质的图像,近年来在国外一些发达国家的渔业管理上已经得到了广泛的应用。分析了DIDSON采集的图像数据,对其进行了图像坐标变换、数据插值以及图像增强等预处理,得到视觉效果较好的声学图像。此外,还对其在渔业应用上的不足进行了分析,给出了扩大声学摄像仪视野的图像拼接方法,以及用于鱼群跟踪、计数的方法,以提高其在渔业上的应用价值。Recently,an acoustic camera called DIDSON（Dual-frequency IDentification SONar） has been widely used in fishery management in some developed countries since it can provide near-video-quality images even in dark or/and turbid water.In this paper,the image data which acquired by DIDSON were analyzed,the pre-processing of the images was conducted,which including coordinates transform,data interpolation and image enhancing,and the acoustic images with improved visual effect were obtained.Then,the defects of the application of DIDSON in fishery were analyzed.In the end,an image mosaicking method to expand the visual field of DIDSON and a method to track and count shoal were proposed to enhancing application value of DIDSON in fishery.

关 键 词：声学摄像仪 DIDSON 图像处理 渔业 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]