원문정보
초록
영어
The test platform plays an important role for the development of intelligent vehicles. In this work, we propose to design and build a miniature test model before building a real scene. The whole test system consists of the simulated traffic environment, the micro-intelligent vehicle and the visual information processing. Firstly, a micro- intelligent vehicle test platform applied to visual information perception is proposed. Secondly, some algorithms about traffic environmental element detection such as lane detection, traffic light and traffic signal recognition and obstacle detection are presented. The experiment validated the proposed approach. On the one hand, the main visual technologies of intelligent vehicles such as line detection, obstacle detection, and traffic signal recognition can be tested in this system; on the other hand, the test result can be used to evaluate and improve the design of the real test platform.
목차
1. Introduction
2. Architecture of Miniature Test Platform
3. Lane Detection and Tracing
3.1. Lane Detection
3.2. Lane Tracking
4. Traffic Light Recognition
5. Traffic Signal Recognition
5.1. Procession in HSV and RGB Color Space
5.2. Edge Detection
5.3. Shape Classification
5.4. Signs Recognition through Multi-layer Perception
6. Obstacle Detection
6.1. Obstacle Detection under Monocular Vision
6.2. Obstacle Detection under Binocular Vision
6.3. Contrast of Monocular Vision and Binocular Vision in Obstacle Detection
7. Conclusions
Acknowledgements
References