Design And Implementation Of An Efficient Hybrid Compressive Sensing Algorithm For Still Images And Videos

Abstract

compression has increasing demands in the field of newlinedigitalization, which needs to be deployed with fast and efficient newlinemethods. In recent times, the world has been inundated with a deluge of newlineinformation and data. The storage and transportation of this information newlineis a challenging task. This has led to the evolution of research into new newlinemethods to represent digital data as compactly as possible. There has newlinebeen great advancement in the data compression field, with specific newlineendeavours for multimedia audio compression, image compression and newlinevideo compression. Most compression ideas are based on the newlinemanipulation of redundancy in the signal. Any data that has embedded newlineredundancy is efficiently expressed in an alternate domain, in which the newlinerequired signal can be sparsely represented. newlineImage compression is a vibrantly changing field with many newlinecompression methods available. Compressed sensing is the mechanism newlinedeveloped in recent times for signal acquisition followed by newlinecompression in one step process. It is a novel scheme that efficiently newlineexploits signal sparsity which is a statistical approach for data newlineacquisition or estimation to sample signals sparsely in transform newlinedomains. The method of compressive sensing replaces conventional newlinesampling and reconstruction with a more general linear measurement newlinescheme with an optimization procedure to collect a subset of signals newlineinside a source at an estimate that is significantly below Nyquist rate. In newlinethis technique, a limited set of data are projected against random vectors newlinex newlineto recover the sparse and compressible signal. A simple linear nonadaptive newlinemeasurement operation which integrates sampling and newlinecompression is adopted in the compressive sensing method. newlineThe aim of this research work is focussed on the investigation newlineof different algorithms on compressive sensing and the implementation newlineof hybrid algorithms on still images and videos. The performance of newlinevarious compressive sensing acquisition and reconstruction methods for newlineimages are estimated and analyzed. The quality of reconstruc