Improved fidelity of triangulation sensor measurements in optical inspection

With the evolution of gear design requirements for new applications, classical gear inspection based on a time-consuming line-oriented tactile measurement must be replaced with a more rapid, areal inspection that can capture complex modern gear modifications. Triangulation-based optical instruments provide a promising path to meet new gear metrology demands with respect to access to the gear flanks and having sufficient speed and accuracy. In triangulation sensor measurement, the image position of a laser line strip on the sensor is analyzed to find the measured geometry. This image of the line on the sensor is calculated through a peak detection algorithm that produces a 'ridge line,' which is the line in the x-y sensor domain with the highest light intensity.
The physics of optical measurement dictates that speckles and scattered light exist during an optical inspection. As a result, when a triangulation sensor is used, the deflection of the scattered light may cause inaccurate peak detection and, therefore, large form deviations in the reconstructed (measured) geometry. In addition, multiple light reflections that influence point calculations from an optical measurement must be detected, eliminated, or remedied. This research provides an improved mathematical approach to ridge line detection in each sensor frame, to detect the peak position of that frame even more accurately. This algorithm is used to measure four reference geometries to evaluate its influence on point clouds from surface measurements when compared to the embedded (OEM) algorithm.
This dissertation offers the improved fidelity of triangulation sensor measurements for optical inspection by developing a novel mathematical approach. It can be used in the future closed-loop control process where the new gear production processes require fast-optical measurement and evaluation processes to trace back from the produced gear geometry to the manufacturing process. This can be achieved by equipping the manufacturing machine with suitable optical measuring devices, an appropriate evaluation strategy, and an inline inspection.