摘要
文章提出了针对二值图像压缩的基于四叉树的分离结构与像素的算法,该算法通过一个阈值来决定对图像像素的合并,在合并过程中生成图像的四叉树结构,并根据结构中节点的取值来决定像素数据的存储,最后形成分离的结构数据与像素线性表。因为在数据分离过程中采用了去冗余处理,该算法对复杂图像的压缩比高于采用JBIG标准所产生的结果;又因为采用了由底至上的合并方案,使该算法有良好的时间和空间复杂性O(N)。由于压缩后的图像含有结构特征,算法可用于建立不同图像之间的联系,同时还适用于对气象、地貌等复杂图像的压缩。
This paper presents a quadtree-based algorithm for compressing binary images by separating their structure and pixel data.A threshold is used to decide whether subimages should be merged into a bigger one to form structure nodes,and the node value is used to decide if the corresponding pixel data should be stored.As pixel data redundancy is eliminated while the quadtree and linear pixel list is constructed,one could get better compression ratio by this algo-rithm than that by a JBIG method when compressing complex images.Since a bottom-to-top merging approach is adopt-ed,the time and spatial complexity of the algorithm is O(N).The algorithm could be used to compress complex images and to form relationships between different images.
出处
《计算机工程与应用》
CSCD
北大核心
2002年第14期85-87,共3页
Computer Engineering and Applications
基金
国家高性能计算基金资助(编号:984054)
