Remote Sensing Small Object Detection Based on Cross-stage Two-branch Feature Aggregation

Authors

Jie Li, College of Information Science and Technology, Qingdao University of Science and Technology, Qingdao 266061, China
Yang Liu, College of Information Science and Technology, Qingdao University of Science and Technology, Qingdao 266061, China
Liang Li, Qingdao Zichai Boyang Diesel Engine Company, Qingdao 266700, China; 3. China Unicom Qingdao Branch, Qingdao 266001, China
Bengan Su, Qingdao Zichai Boyang Diesel Engine Company, Qingdao 266700, China; 3. China Unicom Qingdao Branch, Qingdao 266001, China
Jialong Wei, College of Information Science and Technology, Qingdao University of Science and Technology, Qingdao 266061, China
Guangda Zhou, College of Information Science and Technology, Qingdao University of Science and Technology, Qingdao 266061, China
Yanmin Shi, China Unicom Qingdao Branch, Qingdao 266001, China
Zhen Zhao, College of Information Science and Technology, Qingdao University of Science and Technology, Qingdao 266061, China

Abstract

Abstract: Aiming at YOLOv8's leakage and false detection problems caused by target scale difference and complex background in remote sensing small target detection, this paper proposes a remote sensing image small target detection method based on cross-stage two-branch feature aggregation. The global shared weights in the convolution operator and the context-aware weights of specific tokens in the attention are fused to obtain high-frequency local information and low-frequency global information; the global remote dependencies are captured using a lightweight MLP, and the parallel cross-stage learnable vision center mechanism is designed to capture the information of the local corner regions of the input image; a multidimensional residual attention mechanism is designed to aggregate the output features of two parallel branches to capture pixel-level pairwise relationships as well as cross-channel and cross-space information. The experimental results show that the proposed model achieves 73.8% and 98.1% mAP on DIOR and RSOD datasets respectively, which is 1.3% and 2.1% higher than the current state-of-the-art methods.

Recommended Citation

Li, Jie; Liu, Yang; Li, Liang; Su, Bengan; Wei, Jialong; Zhou, Guangda; Shi, Yanmin; and Zhao, Zhen (2025) "Remote Sensing Small Object Detection Based on Cross-stage Two-branch Feature Aggregation," Journal of System Simulation: Vol. 37: Iss. 4, Article 16.
DOI: 10.16182/j.issn1004731x.joss.23-1526
Available at: https://dc-china-simulation.researchcommons.org/journal/vol37/iss4/16

First Page

1025

Last Page

1040

CLC

TP391

Recommended Citation

Li Jie, Liu Yang, Li Liang, et al. Remote Sensing Small Object Detection Based on Cross-stage Twobranch Feature Aggregation[J]. Journal of System Simulation, 2025, 37(4): 1025-1040.

Corresponding Author

Zhao Zhen

DOI

10.16182/j.issn1004731x.joss.23-1526