Optics and Precision Engineering, Volume. 19, Issue 5, 1082(2011)
Monocular vision system for distance measurement based on feature points
On the basis of target acquisition, recognition and visual blurring, a novel method by using a single camera as a monocular measurement system was presented to overcome the effect of matching of feature points and extraction errors of a single feature point on the measuring results. The principle of monocular vision measurement for target distances based on the feature points was introduced. Firstly, it sets up the mapping relationship between image points and target points according to the pinhole imaging principle, then also describes the area mapping relationship between the object area and the image area of an object by analysis of the object image, and establishes the model of line distance measurement. Finally, the distance measurement between optical center and target was transformed into that between optical center and feature points by processing images and extracting feature points. The proposed measurement principle was verified experimentally. Based on the measuring error increased with the measuring distance, it analyzes the error sources and corrects obtained data. Results show that thelargest error ratio has is 1.68%, as compared with that of 6.72% from the reference 12,both the feasibility and effectiveness of the proposed principle are proved.
Get Citation
Copy Citation Text
HAN Yan-xiang, ZHANG Zhi-sheng, DAI Min. Monocular vision system for distance measurement based on feature points[J]. Optics and Precision Engineering, 2011, 19(5): 1082
Category:
Received: Jul. 7, 2010
Accepted: --
Published Online: Jun. 15, 2011
The Author Email: Yan-xiang HAN (hyx606@126.com)