摘要
针对畜牧储备政策补贴管理中由于养殖场分散和偏远所致畜牧储备数目统计困难的问题,建立了养殖场图像远程采集与目标计数系统。以生猪为实验目标,采用中值滤波和区域生长法进行图像分割,利用基于纹理特征的滤波方法排除颜色灰度相似背景的干扰,实现了远程视频采集的养殖场图像识别计数。该系统的平均识别率在85%左右,能够较准确地识别目标,实现畜牧产品数量的自动计数,可满足一般工程应用中的畜牧产品实时计数的要求。
The number of reserves caused by the scattered and remote farms is difficult to be solved, so a remote image acqui- sition and target counting system, for the livestock farm is established. Taking the pig as the experimental object, the median filter and the region growing method are used to segment the image, and the interference of the color gray background is e- liminated by using the filter based on the texture feature, the image recognition and counting of the remote video capture is realized. The average recognition rate of this system is about 85~, which can accurately identify the target and realize the automatic counting of livestock products. It can meet the requirement of real time counting of livestock products in the gen- eral engineering application.
作者
梁炎森
张天昊
何志毅
LIANG Yansen;ZHANG Tianhao;HE Zhiyi(School of Information and Communication, Guilin University of Electronic Technology, Guilin 541004, China)
出处
《桂林电子科技大学学报》
2017年第6期437-441,共5页
Journal of Guilin University of Electronic Technology
基金
国家自然科学基金(61661016)
关键词
图像识别
多目标识别计数
图像滤波
区域生长法
image recognition
multi target recognition and counting
image filtering
region growing algorithm
