No Reference Medical Image Quality Measurement based on Spread Spectrum and Discrete Wavelet Transform using ROI Processing

摘要

In this paper a new No-Reference (NR) objective quality measurement method based on spread spectrum technique and discrete wavelet transform using ROI processing is proposed. In this method we divide the original image into two separate sub-images called ROI and non-ROI. Region of interest (ROI) is the decision area in a medical image which is very important .This area may indicate a disease and must resulted in a right diagnosis. We use the spread spectrum embedding algorithm to embed a binary mark into DCT transform of non-ROI part of image. This method is useful when guaranteeing a certain level of quality is an important and vital concern. At the receiver side we extract ROI part with least degradation. Then the mark is extracted from non-ROI part and a measure of its degradation is used to estimate the quality of the original image. The performance of our proposed method is evaluated by calculating MSE and PSNR of extracted mark and measuring their correlation with degradation of the whole image. The applications of this work could be compression and storage of images with an acceptable quality level, or compression and transmission over a network for telemedicine applications while preserving an appropriate quality level.