City Research Online

Abstract

This paper deals with a new, reconstruction based, approach of refining a volumetric calibration. The technique is based on a 2D cross-correlation between particle images on the sensor plane with a planar back projection from a tomographic reconstruction in the same sensor plane to determine potential disparities between the initial camera calibration and the measurement. Additive superposition of the correlation maps from different sets or particle images allows reducing the influence of noise and ghost particles such that the systematic errors in the calibration can be corrected. The different sections describe the theory, the principle processing steps and the convergence of the procedure. Furthermore, the concept is proven by simulating the entire process of the measurement chain, with the help of a synthetic comparison. The results show that disparities of over 9 pixels could be corrected to an average of below 0.1 pixels during the refinement steps. Finally, the technique demonstrates it´s potential to measured data, where the numbers of outliers in the raw results are reduced after the volumetric calibration refinement.

Publication Type:	Conference or Workshop Item (Paper)
Publisher Keywords:	PIV processing, Camera calibration, 3D-PIV, TOMO-PIV, Calibration, Calibration Refinement
Subjects:	T Technology > TJ Mechanical engineering and machinery
Departments:	School of Science & Technology > Engineering

Export

Downloads