Optimized JPEG Algorithm Research And Its Application To X-Ray Chest Image Compression

Along with the occurring of new medical imaging technology, such as CT, MRI, and nuclear medicine, the traditional way of analog film-based system gradually loses its dominating position. The filmless medical imaging is proved to be vital. However, the amount of digital medical image captured per-year is large, and increases rapidly every year. The demand for transmission bandwidth and storage space continues to outstrip the capability of available technology, so the stumbling block urges to develop an effective compression technology to achieve a low bit rate in representing digital image while maintaining an acceptable image quality for economical storage and fast transmission. A compression algorithm of X-ray chest image is discussed here, and the principal content of the paper is as follows: Firstly, refer to lots of paper about digital medical image compression, and summarize their strongpoints and limits separately about lossy and lossless coding technology. Secondly, according to the characteristics of X-ray chest image, an optimized JPEG algorithm is proposed. The Huffman code table is built corresponding to every chest image. After analyzing the compressed image, it is quite evident that the quantization is not reversible and leads to the loss of image information. Since data transform and coding technology are near lossless, the quantization loss is the main source of reconstructed image distortion. Therefore, the quantization modified coefficients of every DCT block are calculated through measuring the quantization noise and the regularity of image pixel value of the corresponding block. And then figure out the modified quantization array for dequantization to minimize the reconstructed error. Finally, the above algorithm is implemented in the VC 6.0 platform and thesimulation result confirms its feasibility. The compression ratio is as high as 15:1 while the reconstructed image is isually lossless and satisfies the requirement for diagnosis. In sum, this thesis is on the research of X-ray chest image compression. Image storage and transmission are the chief consideration in the picture archiving and communication system (PACS) and telemedicine. It is important to develop an efficient medical image compression technology and it will have fine application prospect.