移动计算环境中曲线数据实时压缩方法  

Real-time Compression of Curve in Mobile Computing

作　　者：黄跃峰[1,2] 钟耳顺[1,2] 郭会[1,2] 

机构地区：[1]中国科学院地理科学与资源研究所,北京100101 [2]中国科学院研究生院,北京100049

出　　处：《计算机辅助设计与图形学学报》2009年第1期88-93,共6页Journal of Computer-Aided Design & Computer Graphics

摘　　要：移动计算环境中存储、计算和通信等资源受限,为了解决在资源受限环境中实时压缩和解压缩海量曲线数据的问题,提出了基于整形小波变换和FFEP编码的压缩方法.将曲线坐标根据给定容限从浮点数转换为整数,计算其一阶差分后进行整形小波变换;为了加快编码解码速度,提出FFEP编码方案,根据数据的概率分别采用4位、5位、8位及变长码进行编码,对经小波变换后的高频部分有较好的编码效果.由于坐标类型转换、小波变换和FFEP编码主要采用整数位运算及加减法运算并且在原存储空间实现,因此该方法具有较小的时间和空间复杂度.实验结果表明,本文方法能够在当前流行的PDA上将等高线数据实时地压缩到10%以内.Mobile Computing is resource-limited due to its low calculating, storing and transporting capability. A new approach is proposed to compresses and decompresses massive curves in real time under such an environment. Coordinates of curve data are first converted from double float to integer conforming to the given tolerance to reduce redundancy. Afterwards, we calculate their first order difference and perform an integer wavelet transform. ＂FFEP coding＂ is introduced to code and decode the details generated by integer wavelet transform. It adopts codes of 4-bits, 5-bits, 8-bits and variantsize according to the probability of the data. The above processes can be conducted in high efficiency without any additional memory cost. Experimental results show that our approach has a real time performance with a compression ratio below 10%

关 键 词：矢量数据压缩 形变可控 整型小波变换 “FFEP编码” 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]