The performance of the grinding wheel demonstrably affects the machining efficiency and the quality of the workpiece. Therefore, it is essential to evaluate the wear of the wheel and then operate the dressing or replacement in time. The wear procession of the wheel was monitored and evaluated systematically in this paper. A surface grinding experiment was performed by using an alumina wheel to grind the workpiece made of Cr12. The grinding force and the grinding temperature were monitored and measured while the wheel grinds the workpiece. The surface topography of the wheel was also being observed. The distribution of the gray value of pixels in the image of the wheel surface was analyzed by the method of the histogram. Processing of the binary image of the wheel was performed after determining the gray threshold of the gray value. Then, the blockading and the wearing area on the grinding wheel were calculated. Moreover, the relation of the projection area of a single abrasive derived from theory and derived by image recognition was studied. The results of the grinding experiment show that wheel performance degradation occurs when the material removal volume reaches 210 mm3/mm. At this time, the ratio of blockage area on the grinding wheel reaches 13.4%. The percentage of the wearing area is 9.5%. The method of image recognition combined with grinding temperature is workable to realize monitoring and evaluating the wear of wheels on site without unloading them.