Correction of start-up time difference-induced measurement errors of a high-speed binocular stereovision system

In a high-speed stereovision system, high-speed cameras (HSCs) are used to record images and time information, and the measurement accuracy depends on the time precision of the cameras. When two HSCs are used in a binocular stereovision system, the image acquisitions of both cameras are unlikely to start simultaneously. The images that are supposed to be captured at the “same time” may record different states of the test object during an experiment, which will cause measurement errors. In this work, a theoretical model was established to correlate the start-up time difference (STD) and the STD-induced measurement errors of a binocular stereovision system. A simulation of the deflection of a cantilever beam verified the accuracy of the theoretical model. A suggestion of the experimental layout was discussed to reduce the STD-induced measurement errors in one direction, and a method was proposed to correct the STD-induced measurement errors. The effectiveness of the correction method was validated by two experiments, that is, the trajectory of a marker point and the out-of-plane deformation of a bubble. The results showed that the absolute errors of the marker point trajectory were almost negligible and the deformation contours of the bubble were more symmetric after the correction. This correction method should be useful for three-dimensional measurements of physical quantities using a binocular stereovision system.