logo
banner

Journals & Publications

Publications Papers

Papers

A Robust Detection Method of Control Points for Calibration and Measurement With Defocused Images
Oct 30, 2017Author:
PrintText Size A A

Title: A Robust Detection Method of Control Points for Calibration and Measurement With Defocused Images

 Authors: Ding, WD; Liu, XL; Xu, D; Zhang, DP; Zhang, ZT

 Author Full Names: Ding, Wendong; Liu, Xilong; Xu, De; Zhang, Dapeng; Zhang, Zhengtao

 Source: IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT, 66 (10):2725-2735; 10.1109/TIM.2017.2709561 OCT 2017 

Language: English

 Abstract: This paper presents a detection method of control points for the camera calibration and measurement applications, which is robust to defocus. Our method is based on a ground truth, which we call ridge invariance. That is, the positions of broad-brush lines' ridge lines are invariant to defocus blur. First, the ridge invariance is deduced in theory. Then, the methods for ridge point's detection including defocus degree estimation and salience enhancement are deduced. In calibration and measurement experiments, new marks are designed and control points are determined at intersects of the ridge curves. Experiments show that our method can obtain precise calibration and measurement results with images in a large defocus amount range. In the effective defocus amount range, the accuracy of the proposed method almost keeps unchanged to the best values. The proposed method has approximated the same performance as conventional methods at good focus values.

 ISSN: 0018-9456

 eISSN: 1557-9662

 IDS Number: FH3DQ

 Unique ID: WOS:000411025400023

*Click Here to View Full Record