基于视频图像的目标检测技术研究

摘要

本文基于常用运动目标检测方法进行了研究.首先阐述了视频监控系统中常用运动目标检测方法,并对Surendra背景提取算法及改进的帧间差分法进行了说明.运用对称差分法和背景差分法的融合方法进行运动目标检测,实验表明本文运用的方法对运动目标检测具有准确性与实时性.最后得出运用背景差分和帧间差分法的融合检测方法进行运动目标检测,可以得到好的目标检测效果的结论.%The paper study the commonly used moving object detection method.First elaborated the moving target detection method is commonly used in video monitoring system,and Surendra background extraction algorithm and the improved interframe difference method.Using symmetric difference method and background subtraction method of fusion method for moving target detection,experiment shows that this paper uses the method of moving target detection accuracy and real-time performance.Finally by using background subtraction and frame difference method between fusion detection method for moving object detection,can get a good effect on target detection.