计算机科学
突出
人工智能
GSM演进的增强数据速率
计算机视觉
特征提取
特征(语言学)
目标检测
对象(语法)
领域(数学)
干扰(通信)
方向(向量空间)
感知
模式识别(心理学)
计算机网络
哲学
语言学
频道(广播)
几何学
数学
神经科学
纯数学
生物
作者
Xiangyu Zeng,Mingzhu Xu,Yijun Hu,Haoyu Tang,Yupeng Hu,Liqiang Nie
出处
期刊:IEEE Transactions on Geoscience and Remote Sensing
[Institute of Electrical and Electronics Engineers]
日期:2023-01-01
卷期号:61: 1-16
被引量:2
标识
DOI:10.1109/tgrs.2023.3300317
摘要
In recent years, the task of salient object detection in optical remote sensing images (RSI-SOD) has received extensive attention. Benefiting from the development of deep learning, much progress has been made in RSI-SOD field. However, existing methods still face challenges in addressing various issues present in optical RSI, including uncertain numbers of salient objects, cluttered backgrounds, and interference from shadows. To address these challenges, we propose a novel approach, Adaptive Edge-aware Semantic Interaction Network (AESINet) for efficient salient object detection. Specifically, to improve the extraction of complex edge information, we design a Local Detail Aggregation Module (LDAM). This module can adaptively enhance the edge information of salient objects by leveraging our proposed difference perception mechanism. Notably, our difference perception mechanism is a novel edge enhancement method without the supervision of edge groundtruth. Additionally, to accurately locate salient objects of varying numbers and scales, we design a Multi-scale Feature Enhancement Module (MFEM), which effectively captures and utilizes multi-scale information. Moreover, we design the Deep Semantic Interaction Module (DSIM) to identify salient objects amidst cluttered backgrounds and effectively mitigate the interference of shadows. We conduct extensive experiments on three well-established optical RSI datasets and the results demonstrate that our proposed model outperforms 14 state-of-the-art methods. All codes and detection results are available at https://github.com/xumingzhu989/AESINet-TGRS.
科研通智能强力驱动
Strongly Powered by AbleSci AI