论文部分内容阅读
提出一种新的基于小波和低复杂度矢量量化的视频编码算法。利用子带间的相关性，通过同方向跨带矢量中部分低频系数分量间失真的计算代替整个失真的计算，将矢量量化器的计算复杂度降低了７６ ‰；通过最低频子图像的运动矢量去预测其余子图像的位移矢量，极大地降低了运动补偿的复杂度。用Ｍｉｓｓ 序列标准图像作测试，平均压缩比为１６８∶１，平均信噪比为３７．８－ 研究表明该算法具有很低的复杂度和较好的编码性能
A new video coding algorithm based on wavelet and low complexity vector quantization is proposed. By using the correlation between subbands, the computation of the entire distortion is replaced by the calculation of the distortion of some low frequency coefficients in the same direction cross-band vector to reduce the computational complexity of the vector quantizer by 76 ‰. The motion vector of the lowest frequency image To predict the displacement vectors of the remaining sub-images, the complexity of motion compensation is greatly reduced. Using Miss sequence standard images for testing, the average compression ratio is 168: 1, and the average signal to noise ratio is 37.8. Studies have shown that the algorithm has very low complexity and good coding performance