Medical Image Compression By Combining Contourlet Transform With DPCM

Medical image is the important basis for clinical diagnosis. There are a lot of detailinformation which can make an important effect to doctor’s diagnosis. However, there alsoexists a mass of redundant information which makes a big storage content. So, an effectivecompression can not only solve the storage space, but also reduce the transmission time.The paper analyses the characteristics of medical image and makes a detailed analysis onthe existing medical image compression method according to the evaluation of the quality ofcompression standard. The paper mainly makes a research of embedded coding based onwavelet transform, EZW coding and SPIHT coding. To analyze the basic principle and thealgorithm process of the two methods and do experiment simulation according to differentmedical images. At last, we summarize the shortage of the two methods.This paper proposes a new Medical Images Compression method, which combines theContourlet Transform and Huffman encoding with DPCM. DPCM can remove the most ofredundant information between pixels. The original image is predicted by DPCM, we get aprediction image and an error image. Contourlet Transform is not only has characteristics ofmulti-scale, time-frequency localization, but also a high directivity and anisotropy. Besides,Contourlet transform can capture more detail information and overcome the limitations of thewavelet transform decomposition of the direction. The error image is transformed withContourlet transform, then treated with Huffman encoding, so we can get compressionstreaming. Through MATLAB experiment simulation, the results proved this method iseffectively to the medical image compression and can obtain a good reconstructed image. Ithas better feasibility and practicability.