基于小波变换的时空数据压缩方法

Spatio-temporal data compression method based on wavelet transform

介绍了一种基于等时间间隔变化的空间数据的无损压缩算法。考虑到时空数据具有线性变化的规律,本算法采用一维哈尔小波变换,将信号分别通过低通滤波器和高通滤波器,将数据的近似信息储存到低频分量中,细节信息存储到高频分分量中。经小波变换后,小波系数在空间分布上聚集性更强,大量的幅值分布在零附近,即使感知对象异常变化,感知数据存在波动异常,小波变换的多级分解特性可以缓解异常波动对整体数据的影响,保证了高频系数值仍然近似为0。根据小波变换会产生较多幅值较小的小波系数的特点,对小波系数进行最短长度的编码,实现数据压缩的目的。

A lossless compression algorithm of spatial data based on time interval change is carried out. Considering the spatio-temporal data with linear change rule, this algorithm uses a one - dimensional Haar wavelet transform. Will signal has respectively been through the low-pass filter and high - pass filter, the data of approximate information stored in the low frequency component, and detail information is stored to high frequency component. After wavelet transform, the wavelet coefficient of concentration in the spatial distribution is stronger, and a large quantity of amplitude distribution is near the zero. Even if the perception object changes make perception data is abnormal fluctuations, multistage decomposition characteristics of wavelet transform can reduce the abnormal fluctuations impact on the overall data to ensure the high frequency coefficient value is approximate to zero. The wavelet transform can produce smaller amplitude characteristics of wavelet coefficient, and the shortest length of wavelet coefficients coding, and achieve the purpose of data compression.