嵌入式零树小波编码在医学图像压缩中的应用  

Embedded Zero-tree Wavelet Encoding in Medical Image Compression

作　　者：王世刚[1] 游敏娟 

机构地区：[1]泰山医学院放射学院 [2]山东医药技师学院信息中心

出　　处：《中国医学物理学杂志》2010年第4期2024-2026,共3页Chinese Journal of Medical Physics

摘　　要：目的:对嵌入式零树小波编码算法提出改进,实现CT图像压缩。方法:分析嵌入式零树小波编码的原理和算法并探讨小波基的选择,针对嵌入式零树小波编码算法的缺点进行改进,研究设计实现改进算法的程序,进行CT图像压缩。结果:在相同压缩比的情况下,改进的嵌入式零树小波编码算法获得的峰值信噪比比原有算法获得的峰值信噪比高,特别是在高压缩比的情况下,图像的主观质量和峰值信噪比都有了较大的改善。结论:改进的嵌入式零树小波编码算法比原有EZW算法更加优越,能够满足医学图像的传输和诊断要求。Objective： The embedded zero-tree wavelet coding algorithm suggests improvements to achieve CT images compres- sion. Methods ： The basic fundamentals and algorithm of the EZW were analyzed and the choice of wavelets for EZW was re- searched. The EZW was improved to in view of the defect of EZW; a program was written and applied to accomplish the com- pression of CT image. Results： The PSNRs of the improved EZW were higher than those of EZW at the same compression ratios. The image quality and PSNRs got larger melioration especially in high compression ratios.Conclusion： The algorithm of the im- proved EZW is superior to EZW. It can fulfill the requirements of transmission and diagnosis of medical image.

关 键 词：小波变换 图像压缩 嵌入式零树小波 

分 类 号：TP317.4[自动化与计算机技术—计算机软件与理论]