Research On X-ray Image Compression Technology Based On Region Of Interest Coding

X-ray image can directly reflect the internal structure of objects, which is an importantmethod of product quality detection, is widely applied in the oil industrial inspection. It alsoproduces great amount of X-ray image data. The compression on these data facilitates thetransmission and storage, which gradually attracts people’s attention. So, the research on ahigh-efficiency X-ray image compression method has the value of practical application.Based on analysis and research on the current image compression technology, the paperis targeting the X-ray image collected by a detect center of DaQing Oilfield, and aimed at thecharacteristics of X-ray image and the evaluation requirements of the image. It proposes that,based on the integer wavelet transform，combing the embedded image coding and the regionof interest (ROI) coding for the X-ray image compression, and verify the result by theexperiments, the main research contents are as follows:1. Research and analysis the characteristics of X-ray image. Depending on the featuresof high correlation of X-ray image space and rich high-frequency information, it adopts thecorresponding wavelet transformation image coding method, and meanwhile combines thequality requirements of X-ray image, so as to start the research on the integer wavelettransform that can realize the image lossless compression. To select the type of wavelet base,the boundary extension scheme and the wavelet transform level, which makes a foundationfor the high-quality compression of the X-ray image.2. Research on the lossless coding compression scheme of X-ray image based on theinteger wavelet transform. Depending on the characteristics of high correlation of X-rayimage space, it adopts the lossless DPCM coding method to conduct the high-efficiencycoding, and meanwhile combines the SPIHT algorithm and adaptive arithmetic coding. Itproposes the lossless coding compression scheme of X-ray image based on integer wavelettransform, and proves the effectiveness of the scheme through relevant experiments.3. Research on the X-ray image compression method based on the region of interestcoding. The main detection area of the X-ray image is regarded as the region of interest. Theimage segmentation method is combined with the image compression method. And the levelset image segmentation method is applied in the selection of X-ray image region of interest,which solves the problem of inaccuracy of the selection method of traditional region ofinterest. It proposes the X-ray image compression scheme based on the integer wavelettransform and the region of interest coding. The results of the experiments indicate that the scheme effectively improves the compression ratio of the image while it ensures the qualityof X-ray image compression.