碩士 / 東南技術學院 / 防災科技研究所 / 94 / The main objective of this research is to investigate in the feasibility study of passenger safety monitoring using image recognition and laser scanning. Although image recognition technology has been widely used in pedestrian detection for decades, there are disadvantages about image recognition: (1) Limited immunity to lighting conditions and shading. (2) Occlusion problem. (3) Limited camera shooting angle and position. In comparison, laser scanning using 2D or 3D laser scanners developed in few years ago seems to be a perspective tool of replacing image recognition in pedestrian detection because laser beam is immune to lighting condition and data fusion from different scanner seems to be a feasible way of solving occlusion problem. In this research, we provide a summary of an extensive literature review on two pedestrian detection approaches mentioned above and demonstration examples of passenger safety monitoring are described as well.
| Identifer | oai:union.ndltd.org:TW/094TNIOT653007 |
| Date | January 2006 |
| Creators | Chung-Yen Wu, 吳宗彥 |
| Contributors | Lung-Wen Chung, 鍾隆文 |
| Source Sets | National Digital Library of Theses and Dissertations in Taiwan |
| Language | zh-TW |
| Detected Language | English |
| Type | 學位論文 ; thesis |
| Format | 83 |
Page generated in 0.0096 seconds