An online glass medicine bottle defect inspection method based on machine vision

摘要

Quality control is a crucial issue in glass medicine bottle fabrication factories due to the large scale ofproduction. Manual inspection cannot keep up with the quality requirements for glass medicine glass bottles, and automatic glass defect inspection is an increasing trend. An online defect inspection method for glass medicine bottles based on machine vision is presented in this paper. A series of image processing algorithms are set up around the analysis of the glass medicine bottle image and its manufacturing features such as high production speed, low contrast, existence of an interference region existing etc. Firstly, the glass bottle is located in the whole image by projection analysis. Then, the interference regions are removed automatically through the local maximum entropy method. To get a smooth defect image, a multiple scales morphology filtering method is implemented. Finally, the defect is detected through multiple morphology operator fusion. The application of an inspection system based on this method in the Shandong glass bottle factory proved that this inspection method satisfied the requirements of the quality check for medicine bottles with its high effectiveness, accuracy and reliability.