Abstract: In the application of remote sensing image to study the distribution of collapse area, fusion method and quality evaluation method are very important for image analysis results. Five fusion methods, Brovey fusion, Gram-Schmidt fusion, multiplicative fusion, principal component fusion and HPF fusion, are used. The results are evaluated subjectively and objectively based on spectral and texture information respectively. The conclusion is obtained that principal component analysis fusion method is best. The fused SPOT-5 image can better reflect the distribution of collapse in the study area.

Key words: SPOT-5 image, image fusion, quality evaluation

摘要： 利用遥感影像研究塌陷区分布情况的应用中，融合方法和质量评价方法对影像分析结果至关重要。通过Brovey融合、Gram-Schmidt融合、multiplicative融合、主成分融合、HPF融合等五种融合方法对图像进行融合，并分别依据光谱、纹理信息，从主观和客观上对结果进行质量评价，得出主成分分析融合方法最佳的结论，融合后的SPOT-5影像可较好地反映出研究区塌陷分布情况。

关键词: SPOT-5影像, 影像融合, 质量评价