基于低秩矩阵分解的运动目标检测

Moving Objects Detection Based on Low-Rank Matrix Decomposition

运动目标检测是视频监控任务的基础问题之一,针对灰度信息,目标检测存在的阴影识别能力差、检测精度低等问题,提出在HSV颜色空间下基于低秩矩阵分解的运动目标检测算法.首先将获取的RGB图像转为HSV颜色空间分量,分别对H、S、V通道构建低秩观测量,进行低秩矩阵优化分解,分离出不同颜色通道的前景和背景分量;组合H、S、V通道分量的前景图像,得到粗略的运动目标区域;再采用HSV颜色阴影去除去除前景图像中的阴影;最后经噪声去除和空洞的填充,检测得到准确的前景运动目标.实验验证表明,与其它方法相比,能够有效地提高运动目标检测的准确度.

Moving objects detection is one of fundamental tasks of video surveillance. Specific to the poor identification capability and low accuracy to shadow in gray information, this paper proposes a novel moving objects detection method based on the combination of Low-Rank Matrix decomposition and HSV color information. Firstly, we convert the images from RGB space to HSV space, construct observation matrix for H, S, V channels, respectively, and optimize the observation matrix through Low-Rank Matrix decomposition to obtain H, S, V channel’s foreground component and background component; combing H, S, V channels foreground component in order to get roughly moving object district. Secondly, the moving shadow should be detected and eliminated from the foreground image, after combining H, S, V channels component to get the row processing foreground objects and the column processing foreground objects, the row processing foreground image and the column processing foreground image are combined to obtain the moving objects image. Finally, by morphological processing and connectivity detection to eliminate the noise, the accurate foreground moving objects can be obtained. The experimental results demonstrate that the proposed method is much better than others in increasing accuracy of moving objects detection.