Fusion of SPOT and SAR Images for Land Cover Classification

SAR與SPOT遙測資料融合於地表分類

吳孟哲(Meng-Che Wu) ； 陳錕山(Kun-Shan Chen)

《航測及遙測學刊》 12卷1期 (2007/03) Pp. 59-72

https://doi.org/10.6574/JPRS.2007.12(1).5

摘要

本文比較利用主成份分析PCA方法應用於像元階層資料融合技術與Dempster-Shafer evidence theory方法於特徵階層資料融合技術。在像元階層資料融合中，由於合成孔徑雷達的資料具有全偏極特性，在此選取了對植被較為敏感的HV極化合成孔徑雷達資料，與具有光譜特性的光學SPOT資料做資料融合處理以利接下來的地物分類。首先，利用小波轉換技術濾除合成孔徑雷達斑駁雜訊，在接下來融合步驟中，主成分分析出來的第一部分(PC1)是用做完濾除雜訊後的合成孔徑雷達取代，在資料融合後，進行地物分類是採用最大似然法來分類融合影像。在特徵溶合中，利用全偏極雷達資料的極化特性結合SPOT資料的光譜特性，提高分類的精確度。首先使用李式濾波器濾除全偏極雷達資料雜訊，接下來同樣是使用採用最大似然法來分類融合影像，(不同的在於全偏極雷達影像使用Wishart機率分布，在光學影像採用multivariateGaussian機率分布)將每個類別中每個像元屬於某個類別的機率值計算出來，再利用Dempster-Shafer evidence theory來結合這些類別的機率值。最後產生出一張新的分類影像。實驗的結果顯示分類的精確度比較於未融合的資料都有明顯提升的效果，也證明了此兩個資料融合方法對於不同資料特性的融合都是很成功的。

關鍵字

像元等級影像融合 ； 特徵等級影像融合 ； Wishart distribution

並列摘要

In this paper, we compared image fusion of optical and radar images by pixel-level image fusion based on Principle Component Analysis (PCA), and feature-level image fusion based on Dempster-Shafer evidence theory. We combined the HV polarization information from SAR and spectral characteristic from SPOT images in an effort to enhance land cover classification. Before the fusion process, wavelet transform was first applied to denoise the SAR image which suffers from speckle contamination due to coherent process. PCA method is then used to fuse the SPOT and SAR images. In so doing, the PC-1 component is replaced by SAR image (approximation image, after wavelet transform) and then the inverse transform is followed. At last, the maximum likelihood classifier was used for both SPOT-XS images and fusion images. In feature-level case, fully polarization information from SAR is used to combine with spectral characteristic from SPOT images, mainly to enhance land cover classification as well. Speckle noise was removed by Lee filter, followed by the maximum likelihood classifier based on different distribution was used for SAR and SPOT images (Based on Wishart distribution and multivariate Gaussian distribution respectively), to extract the conditional probability of each pixel for each class. Finally, Dempster-Shafer evidence theory is then applied, to combine the classified results of SAR and SPOT data. Experimental results show that the classification accuracy is dramatically improved by effective image fusion of SPOT and SAR data. Excellent results were obtained by the proposed method.

並列關鍵字

Pixel-level image fusion ； feature-level image fusion ； Wishart distribution