Image Compression Algorithm Based On Preprocessing

As computers process more and more images and multimedia information,people to the information processing and communication requirements are higher.In order to save more resources,better compression of images has become the current research hotspot.Because the data volume of the image is very large,it is very difficult to store and transmit.Therefore,it is necessary to achieve high bit quality for the image to be transmitted at a relatively small bit rate.This technique is an image compression coding technique.Therefore,this thesis first briefly introduces the history of the JPEG standard algorithm,as well as its basic principles,classification and evaluation criteria.Among them,the specific process and method of JPEG compression coding are introduced as the key introduction.The detailed principles of discrete cosine transform,quantization,entropy coding,etc.,and the functional analysis of each module are also introduced in detail.Secondly,the organization structure of JPEG2000 standard and its coding and decoding process are studied.Its complete coding process includes preprocessing,wavelet transform,quantization and entropy coding.JPEG2000 also offers other new features such as support for both lossless compression and lossy compression,support for progressive transmission of signal-to-noise ratio and resolution,support for region of interest coding,and good fault tolerance and open architecture.characteristic.Finally a preprocessing method of histogram sparsity is proposed based on the zero point of histogram.And it is combined with the preprocessing method of histogram packing and lossless compression of images using common lossless compression methods,such as JPEG2000,ZIP and RAR.The results show that the histogram sparsity is above 0.4,the image can obtain higher compression ratio by using the preprocessing method of histogram packing.And a preprocessing based data compression method is proposed in this paper.Firstly,Improved Gray Scale(IGS)quantization is used to make the image a bit of distorted.Then,the image is compressed combined with histogram packing.The experiment takes JPEG2000 standard algorithm as an example.The results reveal that the proposed preprocessing method can obtain the Peak Signal to Noise Radio(PSNR)1dB-3dB higher than the lossy compression of JPEG2000.The method is an efficient algorithm.