在传统的H.261/3以及MPEG-1/2标准中是采用基于块的压缩算法,可在低比特率运用时,会出现严重的块状效应及蚊式噪声。为了克服上述弱点,近来提出的面向对象的视频编码算法,曾采用时空分割方法将图象分割成若干个具有一致运动的区域。其虽可克服传统编码算法的缺陷,但却需有多次计算复杂的运动模型参数,因而造成巨大的运算负载。本文提出了一种基于二次空间变换的快速时空分割算法(QBFTSS),即在图象空域分割及区域二次空间变换运动模型参数估计的基础上,通过将非线性的参数估计问题线性化,以及运用Househ

Since the traditional H.261/3 and MPEG-1/2 standards employ the block-based compression algorithm, decoded images are suffered from severely blocky artifacts and mosquito effects at low bit rate. To overcome these weakness, object-based video coding approaches are proposed which segment each image into regions of coherent motion.However,the estimation of more complex motion model is required a lot of times and a large amount of computations is,therefore, unavoided.In this paper, a fast temporal-spatio segmentation algorithm is presented using quadric spatial transfirmations motion model.Based on the image spatial segmentation and the estimation of the parameters of quadric spatial transformations motion model, the amount of computation for temporal-spatio segmentation is efficiently decreased through the linearization the estimating for model parameters、Householder transformation and the fast approach forR、zmatrix.