Abstract: 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.