基于改进密码本模型的视频对象检测

提出了自适应块均值码本模型，可以有效地处理动态背景。该方法考虑了块中像素在相邻帧间的时间联系，像素与其相邻像素间的空间联系，而不是认为像素间相互独立，并能够根据背景变化自适应调整图像块的尺寸。改进的码本模型将当前帧分割为大小自适应的非重叠块，当块中变化像素面积在一定阈值范围内时，将该图像块4等分，并用块均值建立粗糙的cache码本；在码本精细化过程中，用Mahalanobis距离计算cache码本问相似度，相似度高的码本融合为一个码本，最终得到精细码本；将改进模型与HSV色彩空间相结合，可以有效地对前景运动目标和背景进行分割。实验表明，改进的码本模型可以有效的处理动态背景，同时在处理速度上可以满足实时前景背景分割的需要。

Video Object Detection Based on Improved Codebook Model

An improved codebook model based on adaptive size of blocks is presented to efficiently segment foreground from dynamic scenes in this paper.The method takes into account the spatial relations between pixels and the adjacent pixels in the blocks rather than considering the pixels independent,and automatically sets the determinating size of the block instead of arbitrarily fixing the size.The model divides video frames into non-overlap blocks with initial size.Unless the area of changing pixels in the block ...