Adaptive Graph Convolutional Network for PolSAR Image Classification

Fang Liu

doi:10.1109/TGRS.2021.3076997

Adaptive Graph Convolutional Network for PolSAR Image Classification

2022-01-01

14

PRE

AI

DOI：10.1109/TGRS.2021.3076997

Original

Original request for help

Save

Abstract

En 中文

Polarimetric synthetic aperture radar (PolSAR) image classification is one of the hottest issues in remote sensing, where studies on pixel-level information and relationship are of great significance. In this article, graph convolutional network (GCN) is employed to accomplish this pixel-level task benefiting from its excellent capability in structure exploration and information propagation between different pixels. To reduce the communication burden between various PolSAR pixels and high computational cost for the whole PolSAR image, an adaptive GCN (AdapGCN) consisting of pixel-centered subgraphs is proposed in this article. In the AdapGCN, a data-adaptive kernel and a spatial-adaptive kernel are introduced to, respectively, model data structure and spatial structure for PolSAR image. Moreover, a multiscale learning structure is integrated to further explore complicated relations between pixels. Extensive comparative evaluations validate the superiority of our new AdapGCN model for PolSAR image classification over a wide range of state-of-the-art methods on three challenging benchmarks.

Keywords:

Task analysis

Kernel

Scattering

Earth

Data models

Covariance matrices

Symmetric matrices

Adaptive graph convolutional network (GCN)

information propagation

polarimetric synthetic aperture radar (PolSAR) image

subgraphs

Adaptive Graph Convolutional Network for PolSAR Image Classification

Abstract

AI Summary

Journal

Organization