A New Method for Monitoring Gears Surface Failures Using Enhanced Image Registration Approach

Abstract

In this paper, we present an image registration approach to cope with inter-image illumination changes of arbitrary shape
in order to monitor the development of micro-pitting in transmission gears. Traditional image registration approaches do
not typically account for inter-image illumination variations that negatively affect the geometric registration precision. Given
a set of captured images of gear surface degradation with different exposure times and geometric deformations, the
correlation between the resulting aligned images is compared to a reference one. The presented image registration
approach is used with an online health monitoring system involving the analysis of vibration, acoustic emission and oil
debris to follow the development of micro-pitting in transmission gears. The proposed monitoring system achieves more
registration precision compared to competing systems. This paper experimentally validates the system's capabilities to
detect early gear defects and reliably identify the gradual development of micro-pitting in gears, so that it could be used in
predictive health monitoring (PHM) systems and overcome the disadvantages of the most commonly used methods, such
as gear flank profile scanning, replica sample analysis and conventional image analysis.