Hierarchical fusion convolutional neural networks for SAR image segmentation

• Published in: Pattern recognition letters Vol. 147; pp. 115 - 123
• Main Authors: Jiang, Yinyin, Li, Ming, Zhang, Peng, Tan, Xiaofeng, Song, Wanying
• Format: Journal Article
• Language: English
• Published: Amsterdam Elsevier B.V 01.07.2021; Elsevier Science Ltd
• Subjects: Artificial neural networks; Decision theory; Dempster-Shafer evidential theory; Feature extraction; Hierarchical fusion convolutional neural networks; Homogeneity; Image processing; Image segmentation; Neural networks; Radar imaging; Receptive field; Synthetic aperture radar; Hierarchical fusion convolutional neural networks; Image segmentation; Dempster-Shafer evidential theory; Synthetic aperture radar
• ISSN: 0167-8655; 1872-7344
• DOI: 10.1016/j.patrec.2021.04.005

•We propose a HIFCNN model for SAR images segmentation.•HIFCNN model sets several different-sized receptive fields.•HIFCNN model extracts hierarchical features.•HIFCNN model integrates hierarchical features based on the Dempster-Shafer theory.•HIFCNN model is robust against speckle and preserves image structure well. Convolutional neural network (CNN) has achieved promising results in image segmentation recently. However, for the segmentation of synthetic aperture radar (SAR) images with complicated scene, the single receptive field in CNN has a limited ability to effectively capture structural and regional information at the same time. In this paper, we propose a hierarchical fusion CNN (HIFCNN) model for SAR image segmentation. At each convolutional layer, HIFCNN sets several different-sized receptive fields, and thus extracts hierarchical features. Concretely, the larger-sized receptive field captures regional information and is robust against speckle, while the smaller one preserves the structural information well. Then, based on the Dempster-Shafer evidential theory, the proposed hierarchical network, HIFCNN, implements a decision-level fusion to integrate these hierarchical features. In this way, the structural and regional information can be accurately captured by different receptive fields, which is beneficial for edge location, structure preservation and region homogeneity in SAR image segmentation. The effectiveness of HIFCNN model is demonstrated by the application to the segmentation of the simulated images and real SAR images.