小波变换在医学图像压缩中的应用

The Study of Medical Image Compression Based on Wavelet Transform

目的:探讨利用小波变换进行医学图像压缩的方法。方法:通过小波变换对图像进行时频局部化分析,将图像分解到多个尺度上,进行多分辨分析。然后对变换后的子图像的小波系数特点进行了分析,讨论了其适用于图像压缩编码的特性和优势。嵌入式零树小波图像编码算法是一种有效的图像压缩方法。在分析嵌入式零树小波图像编码算法的基础上,针对传统嵌入零树小波编码方法存在的不足之处,提出了一种改进的零树小波编码算法。结果:在获得较大压缩比的同时能保证医学图像的重建质量,可以较好地满足PACS对医学图像存储和传输的要求。结论:仿真实验表明,本方法是一种有效的医学图像压缩方法。

Objective： To study the method of image compression based on wavelet transform. Methods： The wavelet transform was a time-frequency information analysis method.It had the characteristies of multi-analysis. In this way,according to the subimage difference of characteristic,we could carry on the different treatment.Using the features of wavelet image coefficients, image was efficiency compressed.Embedded Zerotree Wavelet （EZW） Encoding was an efficient method in image compression.In this paper, the method of image compression based on the Embedded Zerotree Wavelet Encoding was realized. Considering the shortages of EZW, an improved EZW algorithm was achieved. Results： The quality of the medical image compression reached the level of transmission and diagnosis of medial image. Conclusion： The method was reliable and effective in medial image compression.