Abstract
Land cover classification has achieved significant advances by employing deep convolutional network (ConvNet) based methods. Following the paradigm of learning deep models, land cover classification is modeled as semantic segmentation of very high resolution remote sensing images. In order to obtain accurate segmentation results, high-level categorical semantics and low-level spatial details should be effectively fused. To this end, we propose a novel bidirectional gird fusion network to aggregate the multilevel features across the ConvNet. Specifically, the proposed model is characterized by a bidirectional fusion architecture, which enriches diversity of feature interaction by encouraging bidirectional information flow. In this way, our model gains mutual benefits between top-down and bottom-up information flows. Moreover, a grid fusion architecture is then followed for further feature refinement in a dense and hierarchical fusion manner. Finally, effective feature upsampling is also critical for the multiple fusion operations. Consequently, a content-aware feature upsampling kernel is incorporated for further improvement. Our whole model consistently achieves significant improvement over state-of-the-art methods on two major datasets, ISPRS and GID.
Highlights
L AND cover classification aims to assign correct semantic category label to Manuscript received June 8, 2020; revised August 1, 2020; accepted September 4, 2020
In this work, we propose a novel content-aware bidirectional grid fusion network (CBGF-Net), which is characterized by a bidirectional grid feature fusion architecture and a content-aware feature upsampling kernel
We show the visualized results of BGF-Net, CBGF-Net, and competing methods on ISPRS and Gaofen image dataset (GID) datasets in Figs. 4 and 5, respectively
Summary
L AND cover classification aims to assign correct semantic category label (building, water, vegetation, or road, etc) to Manuscript received June 8, 2020; revised August 1, 2020; accepted September 4, 2020. Significant advances have been achieved on this problem, especially after the community begins to embrace deep convolutional neural network (ConvNet) [3], [4] based methods. State-of-art ConvNet based methods view the land cover classification task as a semantic segmentation problem [5]–[8]. With remote sensing image as input, the ConvNets aim to directly fit the ground truth in an end-to-end manner [9]
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
More From: IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.