Injection Tube Direction Recognition Based on Boundary Correction and Support Vector Machine

Abstract

There is need to recognize the direction of injection tube, which plays a key role in the automatic assembly systems. A boundary correction and Support Vector Machine (SVM) based direction recognition method is proposed to automatically determine the direction of injection tube. First, the Moore neighborhood is utilized to detect and extract the contour boundary of tube image, which is pre-processed after captured by camera; Then, the contour image is performed by boundary correction and normalization, and the direction of the injection tube is recognized by the trained SVM; Finally, the mechanical arm is employed to adjust the plastic pipe realizing the injection tube automatic assembly system. Experimental results show that the proposed method can effectively and accurately recognize the direction of the injection pipe, and it has been applied into practice production achieving good results.