摘要
在视频序列图像中,视频图像分割的主要目的是要在视频序列中分割出运动实体,背景差法能够在实时性要求很高的情况下从一段视频中将当前帧图像跟背景图像作差,然后提取出运动目标,而对差图像进行二值化是关键,自适应阈值方法通过对差图像的直方图进行分析寻找直方图上的一个点,该点到直方图最高点与最低点之间的线段的距离最大,为求该点,本文提出了一种快速算法,以该点对应的灰度值作为阈值,这样运动目标就能够被精确地提取出来。
In the video image sequences, the main purpose of video image segmentation is to segment the moving object in video image sequences. In the case of high real-time demand, background difference method is will be difference of the current frame image and the background image from a video, and then to extract the moving object, the key is the difference image binarization. Adaptive threshold method to find a point on the histogram base on analysis of the histogram of the difference image, the distance that point to the line segment between high and low of the histogram is the largest, in order to find the point. This paper proposes a fast algorithm at that point the corresponding gray value as a threshold, this can be accurately extract the moving object.
基金
贵州省科技厅项目"基于计算机视觉的交通信号自动识别系统研究"成果之一
项目编号:黔科合J字[2013]2147
贵州省教育厅科研基金资助项目"水库水面漂浮物智能监控系统的研发"成果之一
项目编号:黔教科[2011]039
125计划资助成果之一
关键词
视频分割
背景差法
运动目标
Video Segmentation
Background Subtraction
Moving Object
