基于形状上下文的血吸虫尾蚴图像识别  

Schistosome Cercariae Image Recognition by Shape Context

作　　者：陈宇环[1] 钟剑[1] 谢亦才[1] 

机构地区：[1]赣南师范学院现代教育技术中心,江西赣州341000

出　　处：《赣南师范学院学报》2011年第6期62-65,共4页Journal of Gannan Teachers' College(Social Science(2))

基　　金：国家科技支撑计划项目(2009BAI78B00);江西省赣州市指导性科技计划项目

摘　　要：目前国内外对血吸虫尾蚴的监测通常是采取人工捕捞方法收集含尾蚴的水体,然后通过显微镜对水中的尾蚴进行肉眼观测.本文采用数字图像识别技术,研究基于轮廓关键点集的形状识别分类算法,并将其应用于血吸虫尾蚴的视频检测识别中,很大程度上提高了对尾蚴进行实时监测的工作效率.At present, domestic and international monitoring of sehistosome cercariae artificial method of fishing is usually taken to collect a body of water containing cercariae and cercariae in water through a microscope for visual observation. In this paper, digital image recognition technology is used to study the critical point set based on contour shape recognition classification algorithms, and then it has been applied to video detection and identification of schistosome cercariae. The cercariae significantly improves the real-time moni- toring of the efficiency.

关 键 词：形状上下文 轮廓关键点集 血吸虫尾蚴 视频监测 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]