基于小波变换多分辨率分析特性的遥感影像动态重构的金字塔模型

Image pyramid model dynamically reconstruction based on multi-resolution analysis of wavelet transform

本文针对高分辨率遥感影像在网络传输与发布中的大数据量问题,在分析传统影像金字塔构建方法的基础上,利用小波变换的多分辨率分析特性引入了一种基于影像动态构建的金字塔模型。该方法在客户端利用小波逆变换动态合成各种分辨率级别的影像,与传统方法相比解决了没有利用金字塔各个层次数据之间的相关性带来的约1/3数据增量问题,而且在网络应用中可以形成一种分辨率递增的流式数据传输模式,非常适合遥感数据的网络发布。基于该模型进行了系统实验,实验结果表明该方法与传统方法相比响应效率明显提高。证明了该模型在遥感数据存储和传输方面的可行性与优越性。

This paper provides a new dynamic building method of image pyramid on the basis of multi-resolution analysis of wavelet transform.This method gets images of different resolution by inverse wavelet transformation on the client,resolves the problem of 1/3 increasing data,which is brought by traditional building method,because in traditional building method the relativity between layers of the pyramid is not used.A flowing data transmitting mode of increasing resolution can be formed,which is very suitable for remote sensing data's network application.Systemic experiment result shows that this method's response is faster than traditional methods,which testifies the feasibility and superiority of this method in the storage and transmission of images.