期刊名称：International Journal of Advanced Research in Computer Engineering & Technology (IJARCET)
印刷版ISSN：2278-1323
出版年度：2013
卷号：2
期号：4
页码：1549-1554
出版社：Shri Pannalal Research Institute of Technolgy
摘要：The increment in the sizes of the images by the technological advances accompanies high demand for large capacities, high performance devices, high bandwidths etc. Therefore, image compression techniques are essential to reduce the computational or transmittal costs. Wavelet transform is one of the compression techniques especially used for images and multimedia files. In wavelet transform, approximation and detail coefficients are extracted from the signal by filtering. Both approximation and detail coefficients are re-decomposed up to some level to increase frequency resolution. In this paper, we propose a framework for constructing adaptive wavelet decompositions using the lifting scheme. A major requirement is that perfect reconstruction is possible with better quality. Once coefficients are generated, the best directional window sizes are determined to obtain the best reconstructed image, which can be considered as an optimization task. Artificial Bee Colony algorithm which is a recent and successful optimization tool is used to determine the directional window size to produce the best compressed image in terms of both compression ratio and quality. Then we compare the performance of proposed method with the existing method in term of both PSNR and compression ratio.
