Machine vision-based method for obtaining image of surface of conveyor belt

The invention discloses a machine vision-based method for obtaining an image of the surface of a conveyor belt. The machine vision-based method is used for monitoring an operation state of the conveyor belt on line and belongs to the technical field of photoelectric detection. On one hand, a linear array CCD (Charge-Coupled Device) camera works together with high-brightness linear light sources for acquiring image information of the conveyor belt, and on the other hand, two or more linear light sources are used for providing light sources for the single linear array CCD camera. According to the method, the two or more linear light sources are used for providing the light sources for the single linear array CCD camera, thus the quantity of the cameras can be reduced; and for a conveyor belt which is not wider than 1.4m, the image of the surface of the conveyor belt can be acquired by a single camera, so the method is used for favorably acquiring the image of the surface of the conveyor belt on line. For a conveyor of long-distance conveying, the machine vision-based method provided by the invention is beneficial to multipoint monitoring of the conveyor belt.