Image Compression Algorithm And Its Application Research Based On Transform Domain

The research and application of image compression algorithms is one of the most active areas in the information technology. Although many image coding algorithms has been proposed, they still need to be further studied comparing to limits of classcal information theory. The main reseach content of this paper are image compression algorithms and their application based on transform domain, involving the transformation methods of singular value decomposition, discrete cosine transform, wavelet transform and so on.Image compression algorithm based on decreasing bits coding, the conception of compression cell, and compression operator about a number of consecutive unimportant algorithm were put forward. The basic idea are as follows. First, wavelet coefficients are divided into three situations by adaptive threshold and get three kinds of character encodings. Then, do decreasing bits coding to character encodings. At last, do binary bits encoding by compresion cell. Image compression algorithm based on singular value decomposition was brought up. Divided the image into small blocks, after singular value decomposition, obtained two orthogonal matrixs and a diagonal matrix. Take the front two columns of orthogonal matrixs, then quantization by equal probability, combined with diagonal matrix, obtain the final coding. Image compression algorithm based on discrete cosine transform was put forward. The position and coding length of DCT important coefficients were brought up. Divided the image into small blocks, after discrete cosine transform, then quantization and coding by quantization matrix.Based on above compressing algorithms, self-recovery watermarking algorithm was brought up. A more secure scrambling locational operator and an operator which could detect and locate the tampered image block were put forward based on the concept of anti-tamper radius. The basic idea of the algorithm is as follows. it first combined image block compressed data with eigenvalue of image block and its offset block, then scrambled or encrypted and embeded them into least significant bit of corresponding offset block. When tamper occured within the limits of the anti-tamper radius, it could determine the location of tampered blocks and recover tampered block with good image quality.The experimental results show that the above algorithms achieved good compression effect. Especially the compression algorithm based on decrease bits coding, when achieved same compression effect, its encoding length is significantly less than EZW and its compression effect is more prominent in high compression ratio. When the above compression algorithms used in self-recovery watermark study, in the most anti-tamper radius, they could be recover the the image blocks content in the damaged areas with better quality. Add another eigenvalue into watermarking, not only can detect whether images were tampered, but also be able to accurately judge where has been tampered.