Integrating Fuzzy Inference System, Image Processing, and Quality Control to Detect Defects and Classify Quality Level of Copper Rods

Human-based quality control can be of low accuracy and, in most cases, may reduce the speed of decision-making. Therefore, various automated quality control systems have been developed. In this paper, the design of an expert system for automatic quality control is investigated to increase the accuracy of a control system. The knowledge used in this system is gathered thanks to experts, and the data are obtained by a camera. However, since knowledge is implicit and data are uncertain, fuzzy logic is utilized for this system. The captured images of the product are inputs of the system, and the state of the process (in control or out of the control) is output. The input images may be noisy; therefore, the pre-processing method is applied; then, a fuzzy rulebased system along with FAST (Features from Accelerated Segment Test) is developed and applied to extract the specific features needed for controlling the process. Then, the control chart is applied to identify whether the process is in “in control” state or not. An empirical case study is also presented to show the capabilities of the proposed approach.