摘要
讨论了小波在信号压缩中由于正交离散小波变换缺乏平移不变性 ,会对信号压缩的性能造成不利的影响。为了解决这一问题 ,在讨论正交离散小波变换时对每个变换尺度上均对偶采样和奇采样的结果进行保留 ,从而衍生出了二叉树结构 ,按相应准则可以确定最优分解路线并由此可重构原信号 ,最后给出的仿真结果证实了该变换方式较标准正交离散小波变换在取得较高压缩能力的同时 。
In this paper, it was discussed that the negative effect on the signal compressibility of the wavelet because the orthogonal discrete wavelet transform is lacking translation invariability in signal compression. In order to avoid this defect, the odd and even sample results are both reserved when the orthogonal discrete wavelet transform is discussed. Thus a bintree is introduced. The optimal decomposition route is defined according to this criterion and the original signal can be reconstructed. At last, the simulation result proves that this transform method not only has a better compressibility than the standard orthogonal discrete wavelet transform, but also has a higher SNR.
出处
《湛江海洋大学学报》
CAS
2001年第2期58-62,共5页
Journal of Zhanjiang Ocean University
