Research on Target Distance Detection Technology of Vehicle Assisted driving Based on monocular Vision
Yang Liu, Hai Huang, Jun Guo, Run Yuan, Minna Xia, Yutuo Chen · 2021
ADAS has been applied in autonomous driving widely and is the active field of AI applications. This is a survey of distance measurement with monocular camera to detect objects in front of the vehicle and anti-collision warning to make driving more safe. We present a new distance measurement method that YOLOv5 algorithm is applied to detect things in front of the vehicle and the improve method based on pixel feature scale ratio is used to measure distance. In our approach, we use cubic Bezier curve fitting algorithm to optimize the measurement data to improve the distance measurement accuracy. Experimental results on real driving picture show that the proposed distance measurement is robust and accurate of vehicle distance estimation.