The Algorithms Research For Image Compression Based On 8-bank Optimal Haar Wavelet Transform

The image contains lot of information and lager data, thus thecompression of image data has been developed as a special research area-image coding. How to organize, store, transmit and the restoration ofimage data effectively, in other words, how to explore more efficient andbetter compression rate image coding techniques, has become one of keytasks of modem information processing techniques.Transformation of image data is the core of the Image codingtechnology. From 1974, the Indian scientists had developed the discretecosine transform (DCT), either theory or application has been takinggreat development in DCT. Particularly in the areas of practicalapplications, DCT is the core foundation of modem image compressionand coding technology, its leading position is difficult to replace by othersso far. The research in orthogonal transformation, which is better thanDCT, has become an interesting topic both in theory and practicalapplication. In this essay, the optimal Haar transform is chosen from themulti-advanced wavelet theory to study not only the performance of theoptimal Haar transform in image compression coding, but also atheoretical description of such transformation, system structure and somedeep mathematical nature. First, this paper introduced the necessity of image compression andthe development & research status of image compression technology.Moreover, for the image compression technology, this paper made adetailed analysis and exposition in image compression coding theory andall kinds of algorithm; as a result, we gave out evaluation criteria of theimage compression methods. At present, in practical application field, thispaper made a detailed presentation about the two most popular standardJPEG and JPEG2000.Second, this paper made a systematic analysis and comparison,which focused on K-L transform,DCT and three classic orthogonaltransformation in wavelet transform; especially, a detailed analysis hasbeen made for the 8-bank optimal Haar wavelet transform chosen in thispaper, by introducing the basic concepts and conclusions of themulti-resolution analysis of the multi-wavelet, this paper gave out amethod to construct our wavelet which we studied, this paper alsointroduced the technology which deal with the borders of the wavelets.Finally, based on the 8-bank optimal Haar wavelet, this paper made thecomparison experiments between the 8-bank optimal Haar wavelettransform and the DCT by using the basic JPEG coding systemframework. In this paper, through a large number of analysis and statisticsof the standard testing pictures, we have solved the image's quantizationmatrix problem on the new transformation. We define a new image format on a new compression method; based on finishing the work tomake software of image compression & decompression in Windows style,we completed the image conversion between JEPG image format and theimage format defined in this paper.