Machine vision based defect detection system for oral liquid vial

摘要

During the manufacture of oral liquid, some surface defects may appear on the oral liquid vial, such as vial cap scratch and vial body crack, which can reduce the quality of the product. In this paper, a machine vision based defect detection system for oral liquid vial is presented. To satisfy the detection requirements of vial, the inspection apparatus and imaging structure of vail are presented. Then, the vial is divided into two parts including cap and body for detecting, and we propose a method based on horizontal intercept projection to detect the defects on the vial cap and a method based on the black top-hat transform and multi-features to inspect the crack defects on vial body. The in-line experiments show that the accuracy of the proposed method is higher than 98%, and the execution time is less than 60ms, which indicate that the machine vision based defect detection system designed by us can effectively detect the defects on oral liquid vial.