Lossless Compression Algorithm For Color Images Based On Prediction

The technique of digital image processing means that in order to obtain a desired result to process and convert digital image. Image allows the transmission of information and expression more simple and quick. With the continuous development of information technology, digital image in various fields, medicine, defense, manufacturing, etc. have been widely used. Lossless compression of digital image means the image can be completely reconstructed from the compressed data. In many applications, such as medical imaging, remote sensing image, image archiving, high-precision image analysis, historical documentation, preservation of works of art and so urgent need for high performance lossless compression algorithm. Therefore, this paper mainly focus on lossless compression of color images based on prediction algorithms, and presents two different structures of lossless color image compression algorithm.First, this thesis introduces the theoretical basis of image compression and color space; research several classic lossless compression algorithms based on prediction; analyze the advantages and disadvantages of classic pixel prediction model prediction algorithm, algorithm structures and algorithms.Then, this thesis proposes an improved intra prediction mode decision algorithm. This algorithm makes full use of intra prediction algorithm to predict the direction of the image texture features, to minimize the range of optimal prediction mode decision, and improve the speed of the algorithm. Finally, the experimental results show that this algorithm is faster than the speed of classical intra prediction algorithm, and the prediction accuracy of the algorithm is similar to the classical intra prediction algorithm.After that, this thesis proposes a lossless color image compression algorithm based on intra prediction. The algorithm reduce the texture complexity of the R, B component in RGB color space by simple arithmetic operations, and selects the proposed improved intra prediction mode decision algorithm as the pixel prediction algorithm. Finally, the algorithm using the gradient direction to correct pixel prediction differences, according to the characteristics of intra prediction algorithm. The experimental results show that the algorithm is superior to JPEG-LS algorithm in image compression rate. However, the complexity of the algorithm is still needed to be further optimized.At last, this thesis proposes a lossless color image compression algorithm based on gradient prediction. The algorithm selects YCg Co color model for image compression. GAP gradient prediction algorithm is selected to predict the Y component because the Y components of the image contained much more detail information. Simplified intra prediction algorithm is selected to predict the Cg, Co component because these components contain fewer detail information. Finally, the prediction difference is amended and coding. The experimental results show that the performance of the algorithm for image compression is superior to JPEG-LS algorithm, but the complexity of the algorithm is still higher than the JPEG-LS algorithm.