A LOSSLESS MEDICAL-IMAGES COMPRESSIVE WAY BASED ONLOCAL TEXTURE FEATURES

摘要

In this paper, a new Medical-images losslessrncompression way, based on the texture features of therncurrent pixel of the images, is presented. The gradientrnmodels of the texture are employed to predict the currentrnpixel and effectively remove the redundancy of texturernfeatures of the medical-images. Remaining the simplicityrnof the JPEG-LS algorithm, the new way is efficient andrnadaptive to real-time applications.