摘要
模糊图像复原和图像压缩都是数字图像处理的重要应用领域。与清晰图像相比,模糊图像进行小波分解后,接近于0的小波系数更多,更适合于图像压缩。根据这个特性,提出了一种新的基于虚拟模糊图像复原的小波压缩算法。通过对图像进行虚拟模糊处理和小波分解后,对不同子带图像的小波系数采用不同的量化及编码方式,以提高压缩比和压缩质量。算法的关键在于复原图像过程中的矩阵运算、小波分解级数和量化步长的选择等几个方面。模拟实验分析了模糊尺度、小波变换及压缩质量之间的关系,证明了算法的有效性。
Blurred image restoration and image compression are both important application areas of digital image processing. Compar-ing with sharp image,we find that blurred image has more small wavelet transform coefficients and is more suitable to for compress-ing. With this characteristic,a new wavelet compression algorithm based on image blurring and restoration is presented. By codingthe different wavelet coefficients in different ways,we can get higher compression ratio and improve the compression quality. Further-more,the relationship of blurring scale,wavelet transform and compression quality is discussed. Experimental results demonstrate theeffectiveness of the algorithm.
出处
《微计算机信息》
2010年第29期210-211,216,共3页
Control & Automation
基金
南京工业大学青年教师学术基金资助项目
基金申请人:刘国庆
项目名称:独立成份分析在雷达快速扫描中的应用
基金颁发部门:南京气象雷达开放实验室(BJG200806)
关键词
图像模糊恢复
小波变换
图像压缩
图像编码
image blurring and restoration
wavelet transform
image compression
image coding