The Algorithm Study Of Lossless Image Compression Based On Wavelet Transform

In the information explosion era, the rapid growth of information expressed in theform of image brings huge challenges for the limited bandwidth and limited computerstorage capacity. Therefore, in order to solve the contradictions among the vast amountsof information and the limited network bandwidth and computer storage resources, theimage data must be compressed. By removing the redundancy of the image, image canbe expressed by a minimal amount of data to achieve the aim of image compression.Wavelet transform is widely used in image compression, because of its goodtime-frequency localization performance, energy compaction, and strong similarityamong different wavelet sub-bands on the same direction after wavelet transform. Basedon this, the objective of this paper is to intensively investigate the lossless imagecompression method based on wavelet transform.Firstly, the basics of image compression are introduced, including the necessity andpossibility of image compression, the classification of image compression, thecommonly used methods of image compression, the assessment criteria of imagecompression, and the international standard of static image compression.Secondly, wavelet transform and the basic idea of image compression based onwavelet transform and the key factors which affect wavelet image compression arediscussed. The features of energy distribution and the distribution regularities of waveletcoefficient after wavelet transform are analyzed based on experimental data, which canprovide guidance for improving SPIHT algorithm.And then, an in-depth study of encoding principle and realization process of theEZW algorithm and the SPIHT algorithm are made. By analyzing the disadvantages ofSPIHT, a method of lossless image compression based on the improved SPIHT isproposed, which mainly includes the following aspects:①predictive coding is used inlow frequency sub-band and the predictive coding streaming is encoded in Huffmancoding;②the way of set partitioning in the last two scans are changed during the codingof high frequency sub-bands, and the SPIHT coding streaming is further encoded inarithmetic coding. Taking the standard gray-scale images as example, the improvedSPIHT is used to realize lossless image compression and is compared with thetraditional SPIHT algorithm. The experimental results indicate that the improved SPIHTalgorithm reduces the compression bit-rate for different images, which shows that the improved SPIHT algorithm is effective.Finally, the advantages and disadvantages of the general displacement method andthe maximal displacement in JPEG2000are discussed, and a new kind of lossless imagecompression which supports region of interest coding priority is proposed. Taking thestandard gray-scale images as example, the improved SPIHT algorithm is used torealize lossless image compression and is compared with SPIHT algorithm, both ofthem support region of interest coding. The experimental results indicate that theimproved SPIHT algorithm which supports region of interest coding, reduces thecompression bit-rate for different images, which shows that the improved algorithm iseffective.