검색 상세

실내 환경에서 RGB-D 센서를 이용한 표면 그래프 기반의 장소 인식

초록/요약

Place recognition is widely used in the loop closure detection in SLAM. The current approach to place recognition is based on RGB image features. However, indoor environments often do not have rich image features. In particular, walls with no features or repetitive and similar image features in the place degrade the performance of place recognition based on RGB image features. This study presents the place recognition method based on the surface graph. This method clusters the point cloud in the surfaces and recognizes a place through a surface descriptor and a surface graph. The advantage of this approach is that it uses the surfaces as features defined as color and shape, and does not use low-level features such as SIFT and SURF. Another advantage is that the proposed place recognition method uses the surface graph to calculate the similarity by considering the spatial relationship between the surfaces. We have experimented on the data set obtained by the mobile robot equipped with a Kinect sensor in the indoor environment. The experimental results show that the proposed place recognition method based on the surface graph (PRSG) is useful and can be used as a loop closure detector.

more

목차

제 1장 서 론 1
1.1 연구 배경 및 동기 1
1.2 연구 동향 2
1.3 연구의 목표 및 방법 3
제 2장 RGB-D 센서 기반의 표면 추출 5
2.1 표면의 법선벡터 추출 5
2.2 표면 군집화 6
제 3장 장소인식을 위한 표면 서술자 9
3.1 배경 지식 fast point feature histograms 10
3.2 형상과 색상을 활용한 표면 서술자 14
3.3 표면 서술자를 활용한 벽 검출 및 정의 15
3.4 표면 유사도 계산 16
제 4장 그래프 기반의 장소인식 18
4.1 장소에 대한 표면 그래프 정의 18
4.2 장소의 공통 부분그래프 19
4.3 장소 유사도 계산 21
제 5장 실험 및 고찰 23
5.1 그래프 기반의 장소인식 실험 23
5.2 루프결합 검출 실험 26
제 6장 결 론 28

more