Study of 3D splicing method based on control net in stereo vision measurement system

In order to enlarge the measurement range of stereo vision measurement system and realize the accurate measurement for large-scale workpiece, the paper proposes the 3D(three-dimensional) splicing method based on control net. The method constructs the measurement control net on the basis of unit quaternions with the assistant measurement control points set around the measured workpiece, and optimizes the measurement precision of control net with Levenberg-Marquardt optimal algorithm. It transforms the measurement data in the local coordinate systems in different measuring positions to the global space coordinate system, obtains the global measurement data and realizes the 3D splicing of the measurement data. The experiment results show that the proposed method not only enlarges the measurement scope, but also raises the measurement precision for large-scale workpiece.