基于背景色彩识别的磨粒图像分割方法

Wear Particle Image Segmentation Method Based on the Recognition of Background Color

针对铁谱图像背景色彩相对单一的特点,设计了1种基于背景色彩识别的磨粒图像分割方法,采用该方法对轴尖-宝石轴承的润滑油铁谱图像进行图像分割,并用所得磨粒覆盖面积推算轴尖磨损量.通过大量磨粒图像分割实践表明,所设计的方法具有精度高、计算速度快及适应性强等优点.定量磨损分析表明,该摩擦副的磨损进程极为缓慢.

Image segmentation is an important pretreatment in the digital image processing of ferrograph image. The result of common threshold segmentation method is usually inadequate because that there are both dark wear particle and bright wear particle in a ferrograph image. The background color of the ferrograph image is quite simple. Therefore, an image segmentation method based on the recognition of background color was put forward in this paper. Much practice in the ferrograph image segmentation of the wear particles show that the method is good in accuracy and fast in computational speed. A lot of ferrograph images of the pivot-jewel bearing were segmented by the segmentation method based on the recognition of background color. The segmentation resulted in the cover area of the wear particles, then the volume of the wear particles was obtained for the pivot bearing. The wear quantity analysis indicated that the wear rate of the pivot bearing was very small.