Optimization Method for Fractal Image Compression Based on the Maximum Inter-Class Variance Method
Di Xie, Caichun Cen, Yingli Zhao, Jie He, Hongyan Lu, Minglang Chen, Yuan Jiang · 2024
In recent years, with the continuous development of the computer field, the number of various types of image data sets has been increasing day by day. Therefore, the field of image compression [1] has become one of the main development directions to alleviate the problem of data storage space. At the same time, with the progress in the medical field [2] [3], the number of special images such as biological microscopy with strong self-similarity [4] is also gradually increasing. Fractal coding based on image self-similarity and without resolution limitations can alleviate the storage anxiety of medical images such as biological microscopy to a certain extent. However, during the fractal coding process, the global traversal process of the internal range blocks significantly increases the coding time and affects the practical application of fractal coding. Therefore, this paper studies the traversal process of fractal coding. By leveraging the segmentation advantage of the maximum between-class variance method on biological microscopy category images, fractal coding is optimized. The global traversal process is changed to intra-category traversal after adaptive classification of the image. This enables it to improve coding efficiency while minimizing the reduction in decoding quality as much as possible. Experiments are conducted on breast tissue microscopic images and the VOC data set. The experimental results show that the proposed scheme can increase the time used by more than 42% when the peak signal-to-noise ratio drops by less than 1 dB.