Terrain Classification of Polarimetric Synthetic Aperture Radar Images Based on Deep Learning and Conditional Random Field Model
HU Tao,
LI Weihua,
QIN Xianxiang,
WANG Peng,
YU Wangsheng,
LI Jun
Affiliations
HU Tao
①(College of Information and Navigation, Air Force Engineering University, Xi'an 710077, China)②(College of Electronic Countermeasure, National University of Defense Technology, Hefei 230037, China)
LI Weihua
①(College of Information and Navigation, Air Force Engineering University, Xi'an 710077, China)
QIN Xianxiang
①(College of Information and Navigation, Air Force Engineering University, Xi'an 710077, China)
WANG Peng
①(College of Information and Navigation, Air Force Engineering University, Xi'an 710077, China)
YU Wangsheng
①(College of Information and Navigation, Air Force Engineering University, Xi'an 710077, China)
LI Jun
①(College of Information and Navigation, Air Force Engineering University, Xi'an 710077, China)
In recent years, Polarimetric Synthetic Aperture Radar (PolSAR) image classification has been investigated extensively. The traditional PolSAR image terrain classification methods result in a weak feature representation. To overcome this limitation, this study aims to propose a terrain classification method based on deep Convolutional Neural Network (CNN) and Conditional Random Field (CRF). The pre-trained VGG-Net-16 model was used to extract more powerful image features, and then the terrain from the images was classified through the efficient use of multiple features and context information by conditional random fields. The experimental results show that the proposed method can extract more features effectively in comparison with the three methods using traditional classical features and it can also achieve a higher overall accuracy and Kappa coefficient.
