Title :
A Fast Matching Algorithm Based on Adaptive Classification Scheme
Author :
Fan, Ce ; Liu, Pei-hua
Author_Institution :
Fac. of Inf. Sci. & Technol., Guangdong Univ. of Foreign Studies
Abstract :
This paper purposes a novel matching algorithm for image encoding using adaptive classification scheme (ACS) in fractal image compression. It processes based on standard deviation (STD) between range blocks and domain blocks. In this paper, there are two main works, i) the threshold is set to be the ratio of the STD difference and made adaptive ii) we enhance Tong´s STD search algorithm by introducing a domain ACS and classification algorithm, so that domain blocks being matched with similar STD values are located directed. And finally, we present experimental results that show the efficiency of the proposed scheme. It improves the original Tong´s STD algorithm without any loss in the reconstructed image quality, and the encoding time is decreased greatly based on the ACS. It is very to obtain a no search scheme for range-domain match. Also, experimental results shows that images either complex or simple can be encoded in less than 10s
Keywords :
image classification; image coding; image matching; image reconstruction; search problems; ACS; Tong STD search algorithm; adaptive classification scheme; fast matching algorithm; image compression; image encoding; image reconstruction; standard deviation; Classification algorithms; Cybernetics; Electronic mail; Encoding; Fractals; Image coding; Image processing; Image quality; Image reconstruction; Information science; Machine learning; Machine learning algorithms; Partitioning algorithms; Adaptive Classification; Similarity; Standard Deviation;
Conference_Titel :
Machine Learning and Cybernetics, 2006 International Conference on
Conference_Location :
Dalian, China
Print_ISBN :
1-4244-0061-9
DOI :
10.1109/ICMLC.2006.258629
