基于背景重建的运动目标检测算法

针对复杂背景下参考帧的提取问题,提出了一种改进的基于帧间差分的背景提取算法。该算法利用帧间差分将图像序列中的背景像素点提出来,从而确定背景帧;并利用序列图像背景点在时间上呈高斯分布,用“3σ原则”判断背景点是否发生变化来更新背景。实验结果表明,该方法可以有效的提取和更新背景,从而完整准确地检测出运动目标。

Moving Object Detection Algorithm Based on Background Rebuilding

Aimed at the problem that how to attain background frame in complex condition, an improved algorithm based on interframe difference to acquire background is presented. In this algorithm background pixels in the image sequence are picked up by subtracting consecutive frames and therefore background frame is built. The fact that the background pixels of sequence images obey 3a rule of Gaussian distribution was used to update background by judging whether the background pixels are changed. The experimental results demonstrate that this method can be used to extract and update background effectively ,and detect moving object exactly and completely.