论文部分内容阅读
分形编码在图像压缩方面取得了很好的效果 ,同时 ,分形编码也能够用于基于内容的图像检索 .本文提出了一种基于块限制的分形编码算法和匹配策略 ,并将它们用于图像检索 .在我们编码算法中 ,图像会被预先分成互相不重叠的子图像块 ,然后对这些子图像进行独立地分形编码 ,从而获得整幅图像的分形码 .该编码算法能够在很大程度上减少编码时间 .在进行图像间相似性的匹配时 ,我们采用改进的基于九叉树的分配策略 ,从而避免全局地进行分形码的匹配 ,减少了计算量 .实验结果说明 ,我们的编码算法和匹配策略能够比较有效地应用于基于内容的图像检索 ,在计算时间和存储时间上都优于实验中其它两种方法 .
Fractal coding has achieved good results in image compression, meanwhile, fractal coding can also be used for content-based image retrieval.This paper presents a block-based fractal coding algorithm and matching strategy, and they are used for image retrieval In our coding algorithm, the image will be pre-divided into non-overlapping sub-image blocks, and then these sub-images are independently fractal-coded to get the fractal code of the whole image.The coding algorithm can be greatly reduced Coding time.When we do the similarity matching between images, we adopt an improved nine-tree-based allocation strategy to avoid the global fractal code matching and reduce the computational complexity.The experimental results show that our coding algorithm and matching The strategy can be applied to content-based image retrieval more effectively, which is superior to the other two methods in computation time and storage time.