Distance measurement for self-driving cars using stereo camera

• Main Authors: Salman, Yasir Dawood, Ku-Mahamud, Ku Ruhana, Kamioka, Eiji
• Format: Conference or Workshop Item
• Language: English
• Published: 2017
• Subjects: QA75 Electronic computers. Computer science
• Online Access: http://repo.uum.edu.my/22834/1/ICOCI%202017%20235-242.pdf; http://repo.uum.edu.my/22834/; http://icoci.cms.net.my/PROCEEDINGS/2017/Pdf_Version_Chap04e/PID105-235-242e.pdf
• Institution: Universiti Utara Malaysia

Self-driving cars reduce human error and can accomplish various missions to help people in different fields.They have become one of the main interests in automotive research and development, both in the industry and academia. However, many challenges are encountered in dealing with distance measurement and cost, both in equipment and technique.The use of stereo camera to measure the distance of an object is convenient and popular for obstacle avoidance and navigation of autonomous vehicles.The calculation of distance considers angular distance, distance between cameras, and the pixel of the image.This study proposes a method that measures object distance based on trigonometry, that is, facing the self-driving car using image processing and stereo vision with high accuracy, low cost, and computational speed.The method achieves a high distance measuring accuracy of up to 20 m. It can be implemented in real time computing systems and can determine the safe driving distance between obstacles.