基于HSV空间的彩色监控视频前景提取方法

A Method for Extracting the Foreground of Color Surveillance Video Based on HSV Color Space

提出了一种有效的HSV空间视频序列图像背景去除方法。首先,将彩色视频分解为彩色图像帧序列,然后将图像进行高斯滤波,做平滑处理,再利用surendra背景更新算法动态更新背景图像。将当前帧像素矩阵与背景矩阵做差分,通过比较差分矩阵来确定当前帧中前景的区域。实验结果表明,此方法成功地减除了彩色帧图像的背景,较好地保留了前景的颜色,为以后的图像分析打好了基础。

Proposed an effective algorithm based on HSV color space to eliminate the background of the video images.In this paper,the color video is divided into color frames sequence,and then the images are processed by Gaussian filter algorithm for smoothing.Then we update the background image dynamically with Surendra algorithm,which is often used in background building.By subtracting the pixel matrix composed by the current frame from the pixel matrix composed by the background image,a difference matrix is generated,through which the foreground of the current frame is obtained.The result shows that it is a effective algorithm to eliminate the background,and extract the foreground with its original color information,which is propitious to following image analysis.