High-reflectivity objects distributed optical satellite image fusion based on NDVI classification

Ratioing method is one type of the most famous fusion methods in remote sensing image fusion domain. Generally, the ratioing method synthesizes a low-resolution panchromatic (Pan) image by adaptive weighted summation of a multispectral (MS) image. Consequently, the accuracy of the weights for low-resolution Pan image synthesis is of great importance. However, in most cases, the optical satellite images contain lots of high-reflectivity objects, such as clouds covered regions, and high-reflectivity buildings. These objects are saturate due to their strong reflectance. The distortion of saturated objects results in the failure of weights calculation, so that causes the color distortion of fused images. To solve the problem, this paper proposes a high-reflectivity objects distributed optical satellite image fusion method based on NDVI classification. First, the NDVI index is employed to classify the pixels of a MS image into high-reflectivity group and normal group, then the pixels in normal group is used to calculate the weighted coefficients, finally the fused image is obtained by ratioing transform. Experimental results on a large number of test images show that the proposed method has good performance on reducing color distortion.