Defects Detection of Pharmaceutical Blister Packaging Based on Shape Template Matching

摘要

In machine vision defects detection of pharmaceutical blister packaging, the traditional defects detection algorithms always do the binary image processing and edge feature extraction and other operations, and then locate the analysis image through these features. These operations are very tedious and inflexible. In this paper, the shape template matching algorithm is applied to pharmaceutical blister packaging defects detection. The algorithm can determine whether there is a positioning region of the sheet or pieces missing through search algorithm to locate the template image and comparing the difference of gray value between the detection region and the template region. This paper packaged the software of pharmaceutical blister packaging machine vision defects detection via the professional image processing library of Halcon and visual programming software Visual C++. Detection case demonstrates the feasibility and effectiveness of detection algorithm and system.