智能监控系统中运动目标检测算法的研究  

Research on Detection of Moving Objects in Intelligent Monitoring and Control System

作　　者：孙永欣[1] 邓广龙[1] 张博文[1] 

机构地区：[1]黑龙江省科学院自动化研究所,黑龙江哈尔滨150090

出　　处：《自动化技术与应用》2016年第3期73-78,共6页Techniques of Automation and Applications

摘　　要：本文主要针对智能监控系统中的目标检测算法做了改进,采用Canny边缘检测算子,结合帧间差分法、背景减除法、颜色偏差法三种方式的改进算法,实现对光扰噪声、阴影噪声和边缘断裂的去除,从而准确提取出了目标物体边缘,实现对目标物体的检测与分割。This article focuses on the intelligent monitoring system in target detection. The part of target detection uses Canny edge detection operator, combines with the inter-frame difference and background subtraction method, to improve the algorithm, to remove the optical interference noise, shadow noise and edge breaking, and to accurately extract the edge of the target object to achieve the target object detection and segmentation.

关 键 词：目标检测 边缘背景查分法 帧间查分改进算法 

分 类 号：TP277.2[自动化与计算机技术—检测技术与自动化装置]