Oral Product Quality Machine Vision Detection Method

China is a great producer of health-care oral liquid and antiviral oral liquid. Withthe improvement of people’s consumption level and concern for healthy living, thedemand for oral liquid increases greatly. The quality and safety of such products havealso attracted more and more attentions. Currently, most manufacturers in China adoptthe method of manual detection to check the quality of oral liquid products, so it isdifficult to meet the requirements for high-speed, high-quality production. In thispaper,10ml calcium&zinc oral liquid was taken as our research subject. An analysiswas made on the technologies related to the visual detection concerning the quality oforal liquid products.The quality inspection system of oral liquid products was de-signed and developed and the automatic detection of their quality was thus realized.Research and design of the equipment in this paper will greatly improve the qualityand output of the oral liquid products and has great social significance and economicbenefits.First of all, the research background and significance of this paper is introduced,and the present development situation and market demand of the medicine visual au-tomatic detection equipment at home and abroad is analyzed. A brief introduction ismade to the relevant machine visual technologies, their application in practice, theexisting problems and orientation of their development. The overall structure of thevisual inspection system for the quality of oral liquid products is analyzed from suchaspects as mechanical transmission system, intelligent control system and visual im-aging technology. The design in foreign matter separation machinery program andsorting device design and the tracking camera principle are also explored. In addition,the types and characteristics of industrial camera, light source, lens and other visualimaging device are also introduced in this paper. According to this, the imaging de-vice and imaging schemes of the oral liquid visual inspection system are selected anddesigned.Secondly, the capping defects that affect the quality of oral liquid products andthe visible foreign matter detection algorithm in the solution are studied. First, thelinear enhancement and basic filtering-denoising algorithm are introduced. The imagefiltering methods with K-nearest neighbor mean filtering method is used for filteringand a very good effect of filtering is produced. The transverse moment projection al- gorithm is adopted for the detection of oral liquid capping defects. The location mapof oral liquid capping defects is provided, while the method for detecting visible for-eign matter in oral liquid products focuses on the ways to locate the central line of thebottle for oral liquid products. Proposed a fast fuzzy C-means clustering algorithmbased on pixel gray information entropy neighborhood to segment oral liquid bottleimage.The process of locating the central line of the bottle for oral liquid productswith the multi-line search method is explained in details, and the finite differencemethod is recommended for detecting the visible foreign matter in oral liquid prod-ucts.Finally, the software development tools and methods used in the development ofthe system are introduced. Moreover, the functions of frame construction for visualdetection main program, parameter configuration, data management and movementcontrol programs are designed and explained; finally, combined with the design re-quirements of the visual detection system in this paper, the application of multi-linetechnology in the design of the system and the communication process between PCand PLC in the communication system are also examined.