图像块自相似特征的快速分形编码算法

为了解决全搜索分形图像编码算法在编码过程中range块和domain块匹配特别耗时问题,定义了每个range块和domain块的自相似特征,由于在自仿射变换下最优匹配块间的自相似特征应该接近,因此,每个range块的最优匹配块搜索范围仅限在与其自相似特征接近的domain块邻域内,变全局搜索为局部搜索.六幅图像的仿真结果表明,它确实能够在PSNR降低0.48d B(其结构相似性SSIM值仅下降0.0015)的情况下,平均耗时仅为全搜索分形编码算法的18.65%左右,而且也优于其他特征算法,所提算法达到了加快编码过程速度的目标.

Fast fractal encoding algorithm based on image block self-similar feature

The full search fractal image algorithm requires a very long encoding time,which is essentially spent on searching for the best-matched block to an input range block in a large domain pool. In order to solve this problem,defining the self-similar features of each range block and domain block,on account of the self-similar feature for an input range block and its best-matched domain block should be approximate based on the self affine transformation, so the self-similar feature is utilized to confine efficiently the search scope to the vicinity of the domain block having the closest self-similar feature to the input range block being encoded, it can exactly avoid the excessive search.Simulation results of six test images showed that the average time of the proposed scheme is only about 18.65% while there is averagely the PSNR decrease of 0.48dB(its the structural similarity decrease of 0.0015), in comparison with the full search fractal algorithm.Moreover, it is better than the other feature algorithm,the proposed algorithm to speed up the encoding process.