节点文献
用于图像压缩的最优双正交小波设计方法
Optimized biorthogonal wavelets design method for image compression
【摘要】 基于图像压缩中小波基选择的多样性,提出了一种用于图像压缩的最优双正交小波构造方法.该方法给出了一种单参双正交小波的设计准则,它在正则性和滤波器长度上对小波滤波器进行平衡,提高了小波滤波器的频带特性.通过编码增益和时频分辨率的均衡不确定度建立了单参量评价函数,对该函数取最大值可获得应用于图像压缩的最优双正交小波.实验表明,该方法实现简单,只需知道图像的自相关系数,同时获得了比其他同类小波更高的峰值信噪比.
【Abstract】 An optimized biorthogonal wavelets design method is presented to meet the various demands for image compression for wavelet bases.First,a design rule for single-parameter biorthogonal wavelets is described,which helps to balance the regularity and filter length of wavelet filters,and thus improves their frequency characteristics.Then a single-parameter evaluation function is established by regulating the coding gain and unbalanced metric of space-frequency resolution.By maximizing the constructed function,the optimized biorthogonal wavelets for image compression can be obtained.Experiments show that the algorithm is easy to realize,only with the self-correlation of the interested image needed.A higher PSNR is obtained compared with other wavelets of the same category.
【Key words】 coding gain; balanced uncertaintied metric; biorthogonal wavelets;
- 【文献出处】 西安电子科技大学学报 ,Journal of Xidian University , 编辑部邮箱 ,2006年06期
- 【分类号】TN911.73
- 【被引频次】1
- 【下载频次】163