Imaged wear debris separation for on-line monitoring using gray level and integrated morphological features

The characteristics of wear debris particles are valuable information sources for machine condition monitoring. A possible approach is to apply ferrography with computer vision techniques. However, when images are captured on-line, it is observed that particles tend to appear agglomerated and an effective image processing method is hence required. A particle extraction procedure is here developed by making use of advances in morphological segmentations. The reliability of particle separation is improved with both transmitted and reflected debris images. Furthermore, an iterative morphological scaling operation, incorporating gray and boundary based segmentation, is included to increase segmentation accuracy. The performance of the proposed method is tested using real-world wear debris images captured from the lubricant return line in a gearbox. Particle characteristics are found to follow closely the Weibull distribution.