不同温度下仔猪群体特征行为检测方法研究  

Piglets Group Characteristics under Different Temperature Behavior Detection Method Research

作　　者：韩益博[1] 王春光[1] 康飞龙[1] 

机构地区：[1]内蒙古农业大学机电工程学院,呼和浩特010018

出　　处：《农机化研究》2017年第5期21-25,共5页Journal of Agricultural Mechanization Research

基　　金："十二五"国家科技支撑计划项目(2014BAD08B05-04)

摘　　要：针对无应激条件下哺乳仔猪群体生活环境的舒适度检测问题,分别采集低温、适温、高温条件下的哺乳仔猪特征图像,分别对其进行预处理,提取出哺乳仔猪群体目标区域。同时,提出一种基于形态学操作和Radon变换的被测目标区域与其最小包围矩形面积比的计算方法,采用最大类间方差聚类法确定3种特征行为的双阈值,以其作为检测哺乳仔猪特征行为的有效手段,实现了无应激条件下哺乳仔猪群体生活环境的舒适度检测。通过对150张特征行为图片进行实验分析,结果显示:该方法的综合识别率高于87%,为智能监控软件在猪舍温度调控方面提供了数据参考。In order to check the lactation piglet group living environment of the comfort level problem, respectively col- lected under the condition of low temperature, suitable temperature, high temperature characteristics of lactation piglet images. And then carry on the pretreatment, respectively to extract the target area for lactation piglet groups. Based on morphological operations and Radon transform a target area to be measured and the calculation method of minimum boun- ding rectangle area ratio, the between-cluster variance clustering method is used to determine three characteristic behavior of double threshold, with its feature detection lactation piglet behavior effective means, to achieve no stress under the condition of lactation piglet groups living environment of the comfort level detection. Analysis by image characteristics of 150 behavior experiment, the result shows that the method of comprehensive recognition rate above 87%, shows that it has good practical value and application prospect, for the intelligent monitoring software provides data reference in piggery temperature regulation.

关 键 词：哺乳仔猪 舒适度检测 RADON变换 最大类间方差聚类法 

分 类 号：S818[农业科学—畜牧学]