Application of machine learning to classify surface marker of screw

Abstract

In recent years, many character recognition methods had proposed for recognizing handwritten or computer characters. However, there is few paper discusses marker recognition on screw. General speaking, screw images suffered from the influences of splotch and reflection. How to classify the marker of the screw is still a challenging problem. Therefore, we applied the machine learning to recognize the markers on surface of the screw. Experimental results shown that the proposed method achieves reasonable classification accuracy.