Block symmetry predictor to improve fractal image compression. +CD

number: 
1985
English
department: 
Degree: 
Imprint: 
Computer Science
Author: 
Ruaa Abdullah Jaber
Supervisor: 
Dr. Loay E. George
year: 
2008

Abstract:

Various compression methods have been proposed to achieve high compression ratios and high image qualities in low computation time. One of these methods is Fractal Image Compression. The basic idea of fractal image compression is the partitioning of input image into nonoverlapping range blocks. For every range block a similar but larger domain block is found. The set of coefficients of mapping the domain blocks to the range block, using affain transform, is recorded as compression data. The compressed image data set is called the Iterated Function System (IFS) mapping set. Decoding process applies the determined IFS transformations on any initial image, and the process is repeated many times till reaching the attractor. In this research work, four IFS coding schmes have been established and tested. The first scheme is the traditional Fractal Image Compression (FIC) method, it is implemented on color images after transforming the (RGB) color components to (YCbCr) components. The compression results led to encoding time=144.02 sec, compression ratio=8.89 and PSNR=33.39. The second scheme uses the FIC method with a predictor based on centralized moment features, this predictor is introduced to predict the type of symmetry operation required to set the domain block in a proper state to best matches the range block. The use of this predictor helps in reducing the number of trials of symmetry mappings from 8 trials to only one symmetry case. The use of predictor had reduced the encoding to approximately 14% in comparison with that of traditional method. The third and fourth scheme implies the use of FIC method enhanced by the use of moment descriptor (order-1) and (order-3), respectively. Either of these descriptors is used to classify the domain and range blocks into classes, each class is assigned by a class index whose value is equal to moments descriptor value. For encoding each range blocks only the domain blocks have similar descriptor values to that for range block will be IFS-matched with it. In these schemes the symmetry predictor, used in the second scheme, had been used to reduce the search about the best available similar domain block. The attained encoding time in both 3rd and 4th scheme is approximately 0.9% of that spend by a traditional scheme.