对图像采样数据进行矢量变换是实现图像矢量编码的重要组成部分.本文针对矢量变换算法中存在着由矢量构成矩阵的维数和变换核矩阵维数不相等的问题,提出了一种新的矢量变换算法并对该算法的解相关能力和能量压缩能力进行了统计分析.仿真实验结果表明该算法能较大地提高计算效率.

Vector transform of sampled image data plays an important role in image vector coding.To counter the drawback that the dimension of matrix composed by vectors is not equal to the one of transform kernel matrix in vector transform algorithm,this paper presents a novel vector transform algorithm.The capabilities of decorrelation and energy compression of the proposed algorithm are statistically analyzed.The experimental results show that the algorithm can greatly increase computation efficiency.