基于融合差分的可变光强自适应目标识别算法

主要研究视频监控系统中运动目标检测算法,提出一种背景差分与帧间差分相融合的方法。该算法通过多次差分以及判决区域的相关运算划定背景区域和运动区域。同时参考相邻帧平均灰度信息更新背景帧以适应光线变化对判断造成的影响。在图像后处理中结合相关形态学算划分最终的运动目标。该算法可实现运动目标的快速准确定位和区域估算,实验表明该算法的时间复杂度和空间复杂度低,效果良好。

A Novel Algorithm for Variable Light-intensity Adaptive Moving-Objects Detection based on Integration Difference

This paper gives a novel algorithm for moving-objects detection based on fusion of background subtraction and inter-frame difference.This algorithm,via multiple related operations of differences and decision regions,distinguishes moving-objects from background.The background frame based on the average gray between adjacent Images is updated to adapt to the changes in light.In post-processing,some Morphological Operators are used to obtain the complete reliable moving-objects.This algorithm could implement quick positioning of the moving object and object region estimation,and the experiment indicates that this algorithm is fairly low in time and space complexity,excellent in result.