EXAMINATION OF DIFFERENT DOUBLE-PLANE CAMERA CALIBRATION STRATEGIES FOR UNDERWATER MOTION ANALYSIS

Abstract

The purpose of this study was to evaluate five different double-plane calibration strategies with various localized control area sizes and degrees of overlapping to identify the best strategy to minimize the object-space reconstruction error in underwater motion analysis. The object-space reconstruction errors (RMS and maximum) and the RMS-to- max ratios were computed from a simulated underwater calibration trial. An imaginary experimental setup based on a theoretical refraction model was used for a series of camera calibrations. Different double-plane calibration strategies based on the same experimental setup generated very different RMS and maximum reconstruction errors. It was concluded from the analysis that large overlapping localized control areas can substantially reduce the reconstruction error in the double-plane method.