Pixel-wise segmentation of SAR imagery using encoder-decoder network and fully-connected CRF

Gao, Fei and He, Yishan and Wang, Jun and Ma, Fei and Yang, Erfu and Hussain, Amir; Ren, Jinchang and Hussain, Amir and Zhao, Huimin and Huang, Kaizhu and Zheng, Jiangbin and Cai, Jun and Chen, Rongjun and Xiao, Yinyin, eds. (2020) Pixel-wise segmentation of SAR imagery using encoder-decoder network and fully-connected CRF. In: Advances in Brain Inspired Cognitive Systems. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) . Springer, CHN, pp. 155-165. ISBN 9783030394301

[thumbnail of Gao-etal-BICS2019-Pixel-wise-segmentation-of-SAR-imagery-using-encoder-decoder-network]
Preview
Text (Gao-etal-BICS2019-Pixel-wise-segmentation-of-SAR-imagery-using-encoder-decoder-network)
Gao_etal_BICS2019_Pixel_wise_segmentation_of_SAR_imagery_using_encoder_decoder_network.pdf
Accepted Author Manuscript

Download (624kB)| Preview

    Abstract

    Synthetic Aperture Radar (SAR) image segmentation is an important step in SAR image interpretation. Common Patch-based methods treat all the pixels within the patch as a single category and do not take the label consistency between neighbor patches into consideration, which makes the segmentation results less accurate. In this paper, we use an encoder-decoder network to conduct pixel-wise segmentation. Then, in order to make full use of the contextual information between patches, we use fully-connected conditional random field to optimize the combined probability map output from encoder-decoder network. The testing results on our SAR data set shows that our method can effectively maintain contextual information of pixels and achieve better segmentation results.

    ORCID iDs

    Gao, Fei, He, Yishan, Wang, Jun, Ma, Fei, Yang, Erfu ORCID logoORCID: https://orcid.org/0000-0003-1813-5950 and Hussain, Amir; Ren, Jinchang, Hussain, Amir, Zhao, Huimin, Huang, Kaizhu, Zheng, Jiangbin, Cai, Jun, Chen, Rongjun and Xiao, Yinyin