Development of Vision-Based Lane Tracking for Autonomous Cart

• Main Authors: Xiu-Bin Wu, 吳修賓
• Other Authors: Kuo-Ching Chiu
• Format: Others
• Language: zh-TW
• Published: 2012
• Online Access: http://ndltd.ncl.edu.tw/handle/epnk6s

碩士 === 國立虎尾科技大學 === 車輛工程系碩士班 === 100 === In this paper, a vision-based lane tracking system is developed for an autonomous cart. The proposed system is composed of a human-machine controlled interface, an embedded image processing system(NI CVS-1456), a microcontroller(PIC18F252), and a motors controlled system. Based on LabVIEW tool, the human-machine controlled interface is developed in a notebook which takes a wireless communication with the embedded image processing system. When tracking the lane markings, the embedded image processing system uses two CCD cameras to obtain the lane images by protocol IEEE1394 for developing a stereo vision. After extracting the lane images, the 3-D distance information between the lane markings and the cart can be calculated by a stereo vision algorithm. Based on this distance information, the error which defined in image plane will be obtained, and can be transmitted from the embedded system to the microcontroller(PIC18f252) by parallel port communication. Using the error signals, the AFSMC algorithm can be derived from the Lyapunov stability theory by the MCU(PIC18f252) to produce two motors control commands(PWM signals) which can drive the cart to track the center of the lane. A comparative analysis with experimental results soundly confirmed that the performance of developed AFSMC is better than those of classical PID controllers.