Orthogonalized Fractal Image Compression Based On Multiple Image Features

With the advent and rapid development of the digital information age,people have higher and higher requirements for image compression performance.Therefore,fractal image compression coding has become a hot research field in recent years because of its novel,high compression,and independent resolution of fast decoding advantages.However,due to its high computational complexity in coding,the inability to effectively balance the relationship between encoding time and reconstructing high-quality images are still major obstacle in practical applications.Therefore,this article focuses on improving the quality of image reconstruction and the speed of image coding.The main research contents are as follows;(1)Aiming at the problem of excessively long encoding time caused by high computational complexity in traditional fractal image compression,an orthogonalized fractal encoding algorithm based on the texture features of the gray-level co-occurrence matrix is proposed.Firstly,from the perspective of feature extraction and image retrieval,the similarity measurement matrix between range block and domain block is established to reduce the codebook,and the global search is transformed into local search by defining a new normalized block as a new gray description feature,the transformation process between blocks is simplified.Secondly,the concept of synchronous orthogonal matching pursuit(somp)sparse decomposition orthogonalization fractal coding is introduced,which transforms the gray matching between blocks into solving the corresponding sparse coefficient matrix,and realizes the matching relationship between a range block and multiple domain blocks.The final experimental results show that the proposed algorithm can significantly reduce the encoding time while maintaining better image reconstruction quality.(2)Through the research of digital image compression theory and fractal algorithm,it is found that the traditional fractal image coding process is particularly time-consuming because each R block needs to find the best matching block in a the massive codebook,in order to improve this drawback,After in-depth study of the similarity between image sub-regions,a fractal image compression algorithm based on the combination of Hu moment invariants and K-means clustering is proposed,which combines the translation,grayscale,scale,and rotation of the invariant moments.Invariance characteristics,based on the extraction of image features,the R block and D block are clustered accordingly to reduce the matching space.Experiments show that the proposed algorithm has achieved better results.(3)Based on the distribution characteristics of wavelet transform coefficients in the transform domain,a fractal image coding algorithm combining wavelet transform and gray centroid feature is proposed.According to the special sparsity and correlation between frequency bands in the process of wavelet transform,the concept of gray centroid feature is introduced into each high-frequency sub image to detect the image block while retaining the low-frequency image,Simulation experiments show that while the algorithm in this chapter is rapidly increasing the encoding speed,the image reconstruction quality maintains a high level.