Sensitivity of camera's height based on coordinate variation of monocular vision ranging model

Using computer vision technology to perceive the distance of obstacles is an important technology in intelligent driving decision-making. Monocular vision ranging technology has a wide application range and reduces the requirement of camera quality, so it is a hot research direction in the field of...

Full description

Saved in:
Bibliographic Details
Published in2021 International Conference on Artificial Intelligence and Electromechanical Automation (AIEA) pp. 10 - 14
Main Authors Gao, Tian, Li, Meian, Xue, Lixia, Chen, Hao, Zhu, Haojie, Bao, Jingwen
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.05.2021
Subjects
Automation
camera calibration
Cameras
Computer vision
Decision making
Distance measurement
error correction
monocular vision ranging
Sensitivity
sensitivity analysis
Visualization
Online AccessGet full text

Cover

More Information
Summary:Using computer vision technology to perceive the distance of obstacles is an important technology in intelligent driving decision-making. Monocular vision ranging technology has a wide application range and reduces the requirement of camera quality, so it is a hot research direction in the field of visual ranging. In practical application, the change of the camera's external parameters will have a certain impact on the ranging accuracy. This paper will study the sensitivity of the monocular vision ranging model based on the coordinate change to the camera height. We propose a conjecture that the product of the model's formula accuracy and the height measurement accuracy is equal to the final ranging accuracy of the model. This conjecture will be preliminarily verified through experiments in this paper, and an error correction mechanism is proposed to reduce the impact of camera height changes on ranging accuracy.
DOI:10.1109/AIEA53260.2021.00010