| 摘 要: 为了准确地对鱼类的品种进行自动识别，本次研究以鳊鱼、鲫鱼、鳙鱼、草鱼四种淡水鱼为实验对象， 提出一种基于图像处理的鱼类自动识别方法。首先对鱼的外轮廓进行鱼体长度统一化处理，再提取并计算鱼体头部轮廓 夹角、鱼体尾柄宽度与尾柄中点到鱼头夹角顶点的长度之比(宽长比)等特征值，最后运用鱼头部轮廓夹角与宽长比这两 个特征对四种淡水鱼进行分类识别。研究结果表明，本文所提供的分类方法能准确地对鳊鱼、鲫鱼、鳙鱼、草鱼四种淡 水鱼的品种进行识别，且识别精度达到94.97%以上。可为鱼类品种的自动识别提供指导，具有较好的实际应用价值。 |
| 关键词: 图像处理；鱼类品种识别；鱼头夹角；宽长比 |
|
中图分类号: TP391.4
文献标识码: A
|
| 基金项目: 三峡库区生态环境教育部工程研究中心开放课题(KF2015-11). |
|
| An Automatic Identification Method for Freshwater Fishes Based on Image Processing —A Case Study of Four Species of Freshwater Fishes |
|
CHEN Congping,WU Qi,WU Zhe,LV Tian1,2
|
1.( 1.Department of Mechanical and Power, China Three Gorges University, Yichang 443002, China;2.
2.Department of Mechanical Engineering, Changzhou University, Changzhou 213164, China)
|
| Abstract: In order to accurately identify fish species,this study takes four species freshwater fishes,including bream,crucian,bighead,grass carp,as the experimental subjects,and proposes an automatic identification method for fish species based on image processing.The first step is to unify the outer contour's length of the fish body,then to calculate the angle of fish's head and the ratio of the width of the fishtail to the length between the middle point of the fishtail and the vertex of head's angle (ratio of width to length).Finally,the angle of fish's head and the ratio of width to length are used to classify the four freshwater fishes.The results show that the method presented in this paper can accurately identify the four freshwater species,and the accuracy is over 94.97%.It can be used as a guide for the automatic identification of fish species with good practical value. |
| Keywords: image processing;fish species identification;head angle;ratio of width to length |
