Enhanced Remote Sensing Instance Segmentation with Feature Fusion

• Published in: IEEE International Geoscience and Remote Sensing Symposium proceedings pp. 7107 - 7110
• Main Authors: Tao, Zhiwei, Li, Yangyang, Hao, Xuanting, Wu, Bin, Shang, Ronghua, Jiao, Licheng
• Format: Conference Proceeding
• Language: English
• Published: IEEE 07.07.2024
• Subjects: Accuracy; Feature fusion; Image edge detection; Instance segmentation; Location awareness; Prediction algorithms; Predictive models; Remote sensing images; Shape
• ISSN: 2153-7003
• DOI: 10.1109/IGARSS53475.2024.10640731

Instance segmentation in the field of remote sensing imagery is recognized as a complex and difficult task. Previous approaches suffer from inadequate feature fusion, insufficient learning of shape information, and lack of segmentation of object edges. To address these challenges, we introduce FEA-Net(Fusion Edge-Aware Instance Segmentation Network), a multiple information fusion model for remote sensing image instance segmentation. Our model makes the predicted instance masks more accurate and can effectively improve the instance segmentation performance of high-resolution remote sensing images. We have evaluated our method on two datasets, NWPU VHR-10, and the iSAID. The experimental results demonstrate the effectiveness of our method, showing strong performance.