基于鲁棒主成分分析的运动目标检测优化算法

针对鲁棒主成分分析(Robust Principal Component Analysis, RPCA)算法中将动态背景误检为运动目标的问题，该文提出一种运动目标检测优化算法。在RPCA算法初步检测出运动目标后，利用动态背景在时间域上满足高斯分布的特性，以及动态背景和运动目标在整个视频流上检出点均值和方差的差异特性，进一步将动态背景和运动目标分离开来。实验结果表明，所提算法能够有效地处理动态背景的问题，并在一定程度上完整检测出运动目标。

Moving Object Detection Optimization Algorithm Based on Robust Principal Component Analysis

Since dynamic background may be erroneously detected as a moving object in the Robust Principal Component Analysis (RPCA) algorithm, a RPCA-based moving object detection optimization algorithm is proposed to improve it. After detected by the RPCA algorithm, the moving object will be separated from dynamic background according to the Gaussian distribution of dynamic background in the time domain and the difference of mean value and variance between dynamic background and moving object in the whole video stream. The results show that the algorithm can deal with dynamic background effectively and detect the moving objects well.