An algorithm of glass defection on-line detection system

This paper based on a hardware platform of glass defection on-line detection system,it mainly discussed the algorithm of glass defection on-line detection system.Kalman filter is proposed to trace the change of background light intensity.Glass defection is recognized according to the grey variation which changes with the time.A morphology method is used to locating the coordinate of glass defection.The image and coordinate of the glass defection is finally delivered to the computer,it is then sorted by the computer.The algorithm of this paper is applied in a real time processing system that is based on FPGA.The accuracy of this glass defection on-line detection system reaches 0.1mm.