Article ID Journal Published Year Pages File Type
528719 Journal of Visual Communication and Image Representation 2016 12 Pages PDF
Abstract

•A self-recalibration method for multi-view structured light systems is proposed.•The method allows simultaneous reconstruction and calibration.•It is not restricted to special equipment, number of devices or specific arrangements.•The proposal achieves the same accuracy obtained by a chessboard pattern calibration.•Due to its versatility, it can adapt to a wide range of scenes.

Structured light 3D scanning from a single viewpoint requires multiple scans and a registration process for a complete description of the scanned object. Instead, using multiple cameras and projectors simultaneously can reduce the scanning time and increase the visibility. However, a precise estimation of the extrinsic parameters of all the components is a time consuming process prone to errors. This paper proposes a method to automatically reconstruct and self-calibrate multi-view structured light systems with an arbitrary number of devices. The experimentation shows that the proposed method is precise and robust, surpassing other current state of the art approaches. The achieved calibration accuracy is similar to that obtained by a traditional chessboard pattern calibration, but being able to adapt to a wider range of situations.

Related Topics
Physical Sciences and Engineering Computer Science Computer Vision and Pattern Recognition
Authors
, , , ,