Arbitrary-oriented object detection algorithm based on cross-layer integrating attention mechanism for remote images

Jia Lu; Zhihong Chen; Yinping Miao; Baoqi Guo; Xinwang Shao; Shuanglong Xu; Huijun Su

doi:10.1117/12.2641358

4 August 2022 Arbitrary-oriented object detection algorithm based on cross-layer integrating attention mechanism for remote images

Jia Lu, Zhihong Chen, Yinping Miao, Baoqi Guo, Xinwang Shao, Shuanglong Xu, Huijun Su

Author Affiliations +

Proceedings Volume 12306, Second International Conference on Digital Signal and Computer Communications (DSCC 2022); 123060Z (2022) https://doi.org/10.1117/12.2641358
Event: Second International Conference on Digital Signal and Computer Communications (DSCC 2022), 2022, Changchun, China

Abstract

Compared with the classical object detection algorithms with horizontal bounding box, such as Yolo and Faster-R-CNN, the oriented object detector in remote sensing images can be more robust. Considering that the challenge of oriented object detection in remote sensing images, we propose a cross-layer feature fusion improvement strategy. Specifically, to obtain multiple layer feature maps for subsequent object detection, the shallow feature map with texture information merged with the deep feature map with semantic features. The spatial attention mechanism is introduced to enhance our algorithm's attention to the non-local information in these feature maps. Extensive experiments on a public dataset, DOTA, demonstrate the effectiveness of our proposed method. Under the large difference of object scale, arbitrary orientation of object, objects with dense arrangement and complex background, the experimental results show that our method has better performance.

Citation Download Citation

Jia Lu, Zhihong Chen, Yinping Miao, Baoqi Guo, Xinwang Shao, Shuanglong Xu, and Huijun Su "Arbitrary-oriented object detection algorithm based on cross-layer integrating attention mechanism for remote images", Proc. SPIE 12306, Second International Conference on Digital Signal and Computer Communications (DSCC 2022), 123060Z (4 August 2022); https://doi.org/10.1117/12.2641358

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available