A Review of Image Processing for Detecting and Analysing Transmission Error in Gears

Abstract

Transmission error (TE) of gear pairs is the main cause of gear vibration and noise. Over the last years, non-contact optical and vision-based techniques have developed in conjunction with present image processing, point cloud analysis, and machine learning to enable direct, high-resolution assessment of tooth profiles and surfaces appropriate for TE estimates and diagnostic monitoring. Acquisition modalities, image processing pipelines, mathematical techniques for TE computation or profile extraction, and validation criteria are all summarised in this paper.