基于视频的水箱内鱼类行为观测系统  

Automated Observation System for Fish Behavior based on Video

作　　者：徐盼麟[1] 韩军[1] 

机构地区：[1]上海海洋大学海洋科学学院,上海201306

出　　处：《湖南农业科学》2011年第10期140-142,共3页Hunan Agricultural Sciences

基　　金：上海海洋大学研究生基金(A-2501-09-0000-87161)

摘　　要：为了从视频中高效地提取鱼类行为数据,将雷达跟踪多机动目标的先进技术用于鱼类行为的研究,提出了一种基于视频的鱼类行为自动观测系统。选取红鼻鱼(Hemigrammus rhodostomus)为研究对象,以单摄像机结合镜面成像的方法进行观察视频拍摄,通过鱼类行为自动观测系统对鱼类行为数据实现自动化提取。在鱼的三维坐标计算阶段得到鱼的三维坐标算法,借助VC编程对鱼类行为进行自动跟踪,实现了实时输出鱼的鱼类行为数据。To extract the data of fish behavior efficiently from the video, the advanced technology of military radar tracking muhiple maneuvering targets was applied in fish behavior research. Thus a new method that an automated observation system for fish behavior based on video was put forward in this article. Hemigrammus rhodostomus was taken as the target fish of this study. Then, by using a single camera and a waterproof-mirror, the data of fish behavior were extracted automatically by fish behavior observation system. In the process of calculating fish 3D coordinates, VC program was used to track the fish behavior automatically. So we can realize the real-time output 3D data of fish behavior.

关 键 词：鱼类行为 多目标跟踪 自动观测 

分 类 号：S917.4[农业科学—水产科学]