Adaptive Lifting Wavelet Transform and Image Compression Via Texture

张军,成礼智,杨海滨,黄魁华
DOI: https://doi.org/10.3724/sp.j.1016.2010.00184
2010-01-01
Chinese Journal of Computers
Abstract:Natural images are rich in texture,while the conventional wavelet transform apply horizontal and vertical filtering,which fails to provide an efficient representation for image directional features.The wavelet can't achieve ideal sparse representation for natural images,which conclude some defects in image compression.This paper proposes a new adaptive DWT via image texture.Partitioning the image based on local content,predicting the texture direction of local image,keeping the orthogonal property,applying directional wavelet filtering,the method is a chieved by lifting structure with in-place operation.The texture is protected in getting subpixel by interpolating along predicted direction.The method can be used to construct a new image coder with the EBCOT that was adopted in JPEG2000,coding the coefficients and directional information respectively.The result of our method is more efficient for representation of natural images.The experiments of compression show that the numerical result and visual effect are improved by comparing to the conventional wavelet transform.
What problem does this paper attempt to address?