http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
유사 대칭적 구조에 강인한 Point Cloud Submap 기반 로봇 위치 인식
조성준(SungJoon Cho),성규현(Kyu Hyun Sung),김준식(Jun-Sik Kim) 대한전자공학회 2023 대한전자공학회 학술대회 Vol.2023 No.6
Recently, there has been active research on place recognition using Lidar data. These studies utilize SLAM technology, allowing robots to create maps in real-time while determining their location. Various methods are being researched for place recognition, ranging from histogram-based approaches to learning-based descriptor methods. However, existing methods are vulnerable to rotational or translational changes. To address these issues, this paper proposes a map-to-map comparison and matching method using optimized paths and global maps based on consecutive scans, which enables more accurate and robust place recognition. The proposed method is experimentally proven to be superior to existing scan-based methods using the KIST L8 underground parking dataset.