One estimation method of road slope and vehicle distance

Aiming at the problem that the intelligent vehicle cannot accurately obtain the vehicle distance information when the front and rear vehicles are located on the road surface with different slopes, this paper proposes a road slope and vehicle distance estimation method based on vision's and lidar's information fusion. Firstly, the road lane line is detected by combining vision detection and lidar positioning ability, and the slope of the lane line is calculated. Then, we judge and analyze the slope ahead according to the change of the lane line slope, estimate the slope value according to the vanishing point coordinates. Finally, an improved distance estimation method adapted to the conditions of horizontal road surface and slope road surface is designed based on the combination of Lidar and camera. The experimental results show that the improved distance estimation model has good performance and can reach the accuracy of more than 95%.

Automated summary

This paper proposes a road slope and vehicle distance estimation method based on the fusion of vision and lidar information, aiming at the problem of inaccurate vehicle distance information when the front and rear vehicles are on road surfaces with different slopes. Lane line detection is performed using vision and lidar, and the slope is calculated. The slope ahead is analyzed based on the change in lane line slope, and the slope value is estimated based on the vanishing point coordinates. An improved distance estimation method is designed for horizontal and slope road surfaces using Lidar and camera. Experimental results show that the improved model achieves an accuracy of over 95%.