Deep Learning-Based Millimeter-Wave Radar Localization for Indoor Low-Visibility Scenarios
學年 115
學期 1
發表日期 2026-09-09
作品名稱 Deep Learning-Based Millimeter-Wave Radar Localization for Indoor Low-Visibility Scenarios
作品名稱（其他語言）
著者 Shih-Hsuan Huang and Yin-Tien Wang
作品所屬單位
出版者
會議名稱 2026 IEEE International Conference on Artificial Intelligence in Engineering and Technology (IEEE-ICAIET)
會議地點 Kota Kinabalu, Malaysia
摘要 This study investigates self-localization and navigation for rescue robots operating in indoor low-visibility environments. We propose a method that combines millimeter-wave radar sensing with deep learning-based feature recognition to provide robust localization under adverse conditions. Unlike cameras and LiDAR, whose performance degrades in smoke-filled or poorly illuminated environments, millimeter-wave radar remains effective despite smoke and airborne particles. A digital feature map is constructed from radar echo energy distributions and used for indoor localization through feature recognition. To meet the real-time requirements of rescue missions, a lightweight neural network is developed to efficiently estimate robot orientation from radar maps. The proposed method will be implemented on a robotic platform and evaluated in both smoke-free and smoke-filled environments. Experimental results will assess localization accuracy, orientation estimation performance, and system robustness, demonstrating the feasibility of the proposed approach for low-visibility rescue scenarios.
關鍵字 Rescue robots; Millimeter-Wave Radar Localization; Deep learning
語言 en
收錄於
會議性質 國際
校內研討會地點 無
研討會時間 20260909~20260911
通訊作者 Yin-Tien Wang
國別 MYS
公開徵稿
出版型式
出處 Proceedings of the 2026 IEEE International Conference on Artificial Intelligence in Engineering and Technology (IEEE-ICAIET)
相關連結

機構典藏連結 ( http://tkuir.lib.tku.edu.tw:8080/dspace/handle/987654321/129986 )