采用图像处理技术对鱼体健康状况监视和预报

Monitoring and predicting the health condition of fish using image processing technology

该文采用两台PC工业控制计算机分别为现场上位机和远程监控机,以多个PLC可编程控制器和多个单片机系统作为下位机,构成计算机集散监控系统。对养殖水体的多环境因子进行自动监测和控制。特别是当鱼类出现不适或死亡时,鱼体会侧翻,腹部白色的区域会暴露,根据其特点,利用数字图像处理技术可以实现对养殖现场中鱼类的健康状况进行实时监视和预报。首先将摄像头拍摄的现场图像灰度化,然后去除各种噪声干扰,得到质量较好的图像,最后对该图像进行数据统计和识别。对如何实现这种图像处理进行了具体描述,结果表明该系统具有良好的实用效果。

The paper analyzes actuality and existing problems of aquiculture in China and presents the distributed control system which uses two personal industrial computers as a spot host computer and a remote supervision and control computer respectively and many S7-224 programmable logic controllers as well as many single chip processors as secondary computers. The system can automatically supervise and control many environmental factors. Especially when fishes are unwell or dead, they will turn side and the white area of their bellies will be seen. According to features then using the technology of digital image processing, situation of fishes' growth in aquafarm can be real-timely supervised and predicted in time. First, the field image shot by video camera was grayed, afterwards, all kinds of noise interference were removed to gained good-quality image, finally the image was statistically calculated and recognized. The result indicates that the system works well.