Multiscale Gaussian Process-Driven Graph Convolutional Neural Network for Polarimetric SAR Image Classification

By focusing on the structure exploration and information propagation from non-Euclidean data space, graph convolutional neural network (GCN), which can extract abundant and discriminative features, has been a valuable topic in polarimetric synthetic aperture radar (PolSAR) image field. However, the existing GCN-based PolSAR classification methods have high computational cost, may easily be prone to over-smoothing or over-fitting, and inadequately learn the polarimetric property. To address these issues, we propose a polarimetric rotation-based multi-scale Gaussian process-driven GCN (MGPGCN) for semi-supervised PolSAR image classification. Firstly, for addressing the over-smoothing and over-fitting problems, the Gaussian process (GP) is introduced into GCN framework, which can fit the underlying feature distribution rather than calculating specific values of weights in conventional GCN. Secondly, we extend the multi-scale layer architecture and design the multi-scale kernel for improving the representation capability and fully leveraging neighborhood information of GCN. Thirdly, to mitigate the effect caused by noise or imaging angle, a superpixel-level polarimetric rotation-based feature enhancement strategy is designed. With this strategy, the characteristic of each terrain type is more salient, and the representation capability of GP kernel can be further improved. Comprehensive experiments on three PolSAR datasets firmly demonstrate that the proposed MGPGCN can achieve better performance compared with some widely-used GCN-based classification methods.