POSITIONING SENSOR BY COMBINING OPTICAL PROJECTION AND PHOTOGRAMMETRY
Analytics
63 views ◎18 downloads ⇓
Abstract
Six spatial parameters, (x, y, z) for translation, and pitch, roll, and yaw for rotation, are used to describe the 3-dimensional position and orientation of a rigid body - the 6 degrees of freedom (DOF). The ability to measure these parameters is required in a diverse range of applications including machine tool metrology, robot calibration, motion control, motion analysis, and reconstructive surgery. However, there are limitations associated with the currently available measurement systems. Shortcomings include some of the following: short dynamic range, limited accuracy, line of sight restrictions, and capital cost. The objective of this dissertation was to develop a new metrology system that overcomes line of sight restrictions, reduces system costs, allows large dynamic range and has the potential to provide high measurement accuracy.The new metrology system proposed in this dissertation is based on a combination of photogrammetry and optical pattern projection. This system has the potential to enable real-time measurement of a small lightweight module's location. The module generates an optical pattern that is observable on the surrounding walls, and photogrammetry is used to measure the absolute coordinates of features in the projected optical pattern with respect to a defined global coordinate system. By combining these absolute coordinates with the known angular information of the optical projection beams, a minimization algorithm can be used to extract the absolute coordinates and angular orientation of the module itself.The feasibility of the proposed metrology system was first proved through preliminary experimental tests. By using a module with a 7×7 dot matrix pattern, experimental agreement of 1 to 5 parts in 103 was obtained by translating the module over 0.9 m and by rotating it through 60°. The proposed metrology system was modeled through numerical simulations and factors affecting the uncertainty of the measurement were investigated. The simulation results demonstrate that optimum design of the projected pattern gives a lower associated measurement uncertainty than is possible by direct photogrammetric measurement with traditional tie points alone. Based on the simulation results, a few improvements have been made to the proposed metrology systems. These improvements include using a module with larger full view angle and larger number of dots, performing angle calibration for the module, using a virtual camera approach to determine the module location and employing multiple coordinates system for large range rotation measurement. With the new proposed virtual camera approach, experimental agreement at the level of 3 parts in 104 was observed for the one dimension translation test. The virtual camera approach is faster than the algorithm and an additional minimization analysis is no longer needed. In addition, the virtual camera approach offers an additional benefit that it is no longer necessary to identify all dots in the pattern and so is more amenable to use in realistic and usually complicated environments. A preliminary rotation test over 120° was conducted by tying three coordinate systems together. It was observed that the absolute values of the angle differences between the measured angle and the encoder reading are smaller than 0.23° for all measurements. It is found that this proposed metrology system has the ability to measure larger angle range (up to 360°) by using multiple coordinate systems. The uncertainty analysis of the proposed system was performed through Monte Carlo simulation and it was demonstrated that the experimental results are consistent with the analysis.