一种消除重建图像边缘振荡的图像压缩方法

针对小波变换进行图像压缩过程中重建图像易引起边缘振荡的问题,提出了一种非延拓消除边缘振荡的方法。该方向将提升算法与传统的非线性ENO小波变换相结合,把高频分量边缘的能量集中到低频分量上,用小波分量模的极大值对跃变点进行奇异性检测,有效地消除了重建图像引起的边缘振荡。仿真实验表明该方法计算简单、存储空间小。

Image Compression Method for Damping Edge Oscillation in Reconstructed Image

In order to solve the problem of edge oscillation caused by reconstructed image in image compression using wavelet transforms, an essential non-oscillatory without extrapolation(ENOWE) was put forward. Through combination of lifting scheme with traditional nonlinear ENO-wavelet transforms, this method made the energy of image edge in high frequency component concentrate on low frequency component. And the singularity detection for the accurate jumps location was conducted by maximization of wavelet component module. Therefore it could eliminate the edge oscillation in reconstructed image. Simulation experiments show that the proposed method is effective and simple for calculation with small storage space.