The Research On Visual Inspection Technique For Foreign Substances On Medicine Filling Production Line

Medicine filling products are important pharmacy of the pharmaceutical industry and have been widely used in clinical. However, due to the production process or environment reasons, some of the medicine filling products may contain scrap glass, fiber, rubber, hair and other foreign substances, which could cause great damage to human bodies. Therefore, China's Pharmacopoeia stipulates that the examination of foreign substances in medicine filling products must be conducted on bottle-by-bottle basis. Currently, manual inspection is the most popular method among domestic manufacturers. Subjectivity, low efficiency and high emission factors of this manual process are the bottleneck of the automatic medicine filling production line. Machine vision technology is gradually used to inspect the foreign substance of the medicine filling products; the new method accesses the real-time image of medicine filling products quickly and accurately and achieves detecting, measuring, analyzing and diagnosing the foreign substances of the medicine filling products. At the same time, the quality of the medicine filling products is ensured and the automation of the medicine filling production line is improved by the way.Firstly, thesis summarizes the present research status about vision inspection technology of the foreign substances for the medicine filling products. The key machine vision imaging technology and the foreign substances characteristics of the medicine filling products are analyzed, then the visual detection system of the medicine filling production line is designed. Because of the lack of the traditional backlight forms, improving the backlight and lighting back at the bottom are designed to obtain clear images of the medicine filling products. Employing mechanical devices for filling bottles of high-speed rotating - stopping, visual inspection system obtains sequential images of the medicine filling products by controlling image acquisition devices and uses the sequential images of movement and static objects in time and space on the bottle to distinguish between differences in the interference and liquid internal foreign bodies and overcomes the interference of bottle.In accordance with the specific requirements of medicine filling products, the thesis carries out in-depth study and exploration about image de-noising, foreign object detection and extraction of foreign bodies and designs the adaptive median filter algorithm and based on wavelet transform and adaptive fuzzy filter algorithm ,which suppress effectively the image noise and retain well the image details. Against the foreign substances detection, the inter-frame difference method and the combination of background subtraction and time difference method are presented; they can effectively exclude variety possible interferences of the bottle and accurately detect the small foreign body targets in the medicine filling products. By the analysis of the common edge detection algorithms, an improved Canny edge detection algorithm is presented, which improves the precision and accuracy of the foreign substances edge detection.Finally, the thesis introduces in detail the experimental platform that inspects the foreign substances, including the mechanical structure, electrical control system, hit a sorting device and software system design and implementation. It has laid a good foundation in order to further research and develop.