Pansharpening via Triplet Attention Network With Information Interaction

Pansharpening aims to obtain high spatial resolution multispectral (MS) images by fusing the and spectral information in low (LR) MS panchromatic (PAN) images. Recently, deep neural network (DNN) based pansharpening methods have been advanced extensively. Although most DNN-based show good performance, it is difficult for them preserve details fused image. In this article, we propose a new method on triplet attention with interaction efficiently enhance First, different mechanisms are designed model feature properties LR PAN Then, complementarity among maps enhanced interaction, which promotes compatibility of features from subnetworks. Finally, utilize graph module capture similarity within maps. According graph, informative selected provide more reconstruction Extensive experiments QuickBird GeoEye-1 satellite datasets that proposed can produce competitive when compared some state-of-the-art methods.