Self-calibration method of 3D measurement system with double cameras and 3-coding cycle triangular gray distribution structured light

Abstract

This paper puts forward a new kind of structured light 3d measurement system self - Calibration method. At first, this paper sets the polar equation between the feature points on the calibration object and their corresponding points on two images. At second, the paper derives the translation and rotation relationship matrixes of the projector and each camera. Afterwards, projecting 3-coding cycle triangular gray distribution structured light patterns on the calibration object. From the images, the coordinates of the projector of the feature points can be acquired. Finally, the estimation matrix of the relationship matrix can be set based on the improved 8-point method. Using the epipolar distance as the objective function, the relationship matrixes will be optimized, and the parameter matrixes of the two cameras and the projector could be obtained. Making use of the calibrated system are carried out for a plane and a plaster statue respectively. The experimental result shows that the maximum absolute measurement error of the plane is less than 2.3mm and the reconstruction plaster statue surface is close to the target plaster statue. It proves that the 3d measurement system and the self-calibration method used by this paper is effective.