Determination of defects of electrolytic zinc-coated Circular Plugs by using Machine Learning Pretreatment

摘要

The present invention relates to a method for determining electrical galvanizing defects of round plugs through machine learning pre-processing. When galvanizing a round plug for a vehicle using electricity, it is automatically determined whether the galvanizing state is defective without an operator's visual observation. It relates to "a method for determining the electrical galvanizing defect of a circular plug through machine learning preprocessing". The technical idea of the present invention is a method for determining the electrical zinc plating defect of a round plug through machine learning pre-processing, (a) extracting a two-dimensional image of the circular plug (b) applying a boundary enhancement method to the extracted 2D image; (c) extracting a plurality of concentric circles by performing a circular Hough transform method; (d) selecting a minimum diameter concentric circle region having a minimum diameter; (e) generating a rectangular shape enclosing the minimum diameter concentric circle area and combining it with the minimum diameter concentric circle area; (f) removing the pixel values of a peripheral area excluding the concentric circle area with the minimum diameter by setting them to 0 (the area between the minimum concentric circle area and the quadrangle); (g) input to the learned deep learning tool to determine whether the electric galvanizing state of the circular plug is defective In the case of implementing the method for determining the electrical galvanizing defect of the round plug through the machine learning preprocessing proposed in the present invention, it is possible to efficiently determine the most important central region in the electro galvanized state of the round plug, so that the entire circular plug image is Compared to the machine learning method, a very efficient result can be obtained.