With the continuous advancement of cutting-edge technology, safety issues during the driving of autonomous vehicles(AV) have become the hottest topic at present. Aiming at the problems of inaccurate positioning of the front vehicle position and difficulty in real-time calculation and decision-making during the driving process of autonomous vehicles. In order to obtain the accurate distance of the vehicle ahead in real time, we propose a vehicle distance estimation method based on monocular vision and deep learning. First, a convolutional neural network(CNN) for the object of this research is constructed, and the vehicle image is put into this network for training. The position and distance of the license plate are detected from the original image and combined with the principle of geometric vision ranging to realize the positioning and distance of the vehicle in front measurement. The experimental results show that the proposed method can effectively recognize the license plate and detect the distance of the vehicle ahead quickly and accurately, effectively solving the difficult problem of traditional visual ranging and large distance measurement errors.
Vehicle Distance Estimation Based on Monocular Vision and CNN
01.09.2021
1541127 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Vehicle distance detection based on monocular vision
IEEE | 2016
|PROBABILISTIC DISTANCE ESTIMATION FOR VEHICLE TRACKING APPLICATION IN MONOCULAR VISION
British Library Conference Proceedings | 2016
|Inter-vehicle distance estimation considering camera attitude angles based on monocular vision
SAGE Publications | 2021
|Sociai Distance Measuring Based on Monocular Vision
British Library Conference Proceedings | 2022
|