The Research Of Obstacle Avoidance For The Intelligent Autonomous Vehicles

• Main Authors: Ping-Che Chiang, 江秉哲
• Other Authors: Ying-Jeng Wu
• Format: Others
• Language: zh-TW
• Published: 2016
• Online Access: http://ndltd.ncl.edu.tw/handle/04829259548707642653

碩士 === 國立雲林科技大學 === 機械工程系 === 104 === The thesis presents a method for the navigation and obstacle avoidance with the Computer Vision and Light Detection and Ranging sensor. With CV, vehicle can Detect the navigation. With a LiDAR sensor, the environment could also be perceived by the vehicle. The thesis has developed a path-planning algorithm which integrated all the data acquired by these sensors, in order to avoid obstacles in the environment and return the Origin cruising-path. The thesis applied the image process library Emgu CV, which is cost-free to conduct an image analysis and pre-processing. YunRacer use Gray image at first ,and use Canny edge detection ,then map the area of the lane to Bird’s eyes image with IPM. Finally , utilize the Hough Transform to take the information of lane and determine the cruising path. When the vehicle obstruct by obstacles, avoid obstacles by VFH+ algorithm with the data of LiDAR. Finally, The vehicle return the origin cruising-path after obstacle avoidance.