2D DWT lifting image compression scheme for error tolerant applications
Nayna Vijaykumar Bhosale, Sudhir S. Kanade · 2017
Now a day's multimedia and portable devices use high graphics, which is based on images and videos. Due to a high quality of images and video, there is need of compression unit which will reduce the size of raw material & make useful for portable devices. Every portable device is used for as an application of real time transmission and reception of data and in the current stage, every portable and multimedia devices are battery operated. Due to heavy graphics and high computation unit, there are issues of heavy energy consumption. For reduction of that energy consumption, there is need of fast and low power architecture which will make justification with SPPA (Speed, Power, Area, and Accuracy). According to error tolerant analysis for Image/Video, there is 5-10% error which is tolerable by the human eye. So according to 5-10% error margin, there is no need of accurate architecture because in accurate architecture we will lose the efficiency of SPAA metrics. So, we can easily replace accurate architecture with approximate architecture for error tolerant applications. According to approximation concept, if we will add some kind of error in our system which will reduces the time and energy complexity, with small degradation is in image quality which is tolerable by the human eye. In this work, we have proposed a novel architecture of 2D discrete wavelet transform by using of approximation concept for error tolerant applications. Here, we have implemented the algorithm by Matlab. For image quality analysis we have performed two kind of analysis which is objective and subjective. In subjective analysis, we have performed server base analysis in terms of visual quality. Where as in objective, we have used some existing image/video parameters, which are PSNR, SSIM, FSIM, RFSIM, GMSD, Correlation, and Absolute Similarity.