Multitask Learning-Based for SAR Image Superpixel Generation

Remote Sensing

Deliang Xiang¹  Wenbo Jing²  Jianda Cheng³  Jiafei Liu³  Qingsong Wang⁴ 

[1] Beijing Advanced Innovation Center for Soft Matter Science and Engineering, Beijing University of Chemical Technology, Beijing 100029, China;College of Electronic Science, National University of Defense Technology, Changsha 410073, China;College of Information Science and Technology, Beijing University of Chemical Technology, Beijing 100029, China;School of Electronics and Communication Engineering, Sun Yat-sen University, Shenzhen 518107, China;

关键词: multitask learning;    SAR image superpixel generation;    high-dimensional feature space;    pixel-superpixel soft assignment;   

DOI  :  10.3390/rs14040899

来源: DOAJ

【 摘 要 】

Most of the existing synthetic aperture radar (SAR) image superpixel generation methods are designed based on the raw SAR images or artificially designed features. However, such methods have the following limitations: (1) SAR images are severely affected by speckle noise, resulting in unstable pixel distance estimation. (2) Artificially designed features cannot be well-adapted to complex SAR image scenes, such as the building regions. Aiming to overcome these shortcomings, we propose a multitask learning-based superpixel generation network (ML-SGN) for SAR images. ML-SGN firstly utilizes a multitask feature extractor to extract deep features, and constructs a high-dimensional feature space containing intensity information, deep semantic informantion, and spatial information. Then, we define an effective pixel distance measure based on the high-dimensional feature space. In addition, we design a differentiable soft assignment operation instead of the non-differentiable nearest neighbor operation, so that the differentiable Simple Linear Iterative Clustering (SLIC) and multitask feature extractor can be combined into an end-to-end superpixel generation network. Comprehensive evaluations are performed on two real SAR images with different bands, which demonstrate that our proposed method outperforms other state-of-the-art methods.

【 授权许可】

Unknown