Research On Visual Inspection Of Special-shape Transparent Softgel

Softgels are commonly found in pharmaceuticals or health foods,which are capsules made by sealing liquid medicaments in a soft capsule.As the softgel has a beautiful appearance,accurate dose,easy to carry and easy to use,it has become one of the most widely used pharmaceutical.At present,most of the pharmaceutical manufacturers are still using the artificial visual observation method to identify the defective capsules.This method is not only inefficient,but also cause a large number of error detection and missing detection,because the inspectors have visual fatigue and human eye has recognition limit for fine objects.Since the softgel is irregular and its defects are not obvious,the traditional machine vision defect detection technology is not suitable for the detection of special-shaped transparent softgel.In this paper,after a full analysis on the related research and its limitations,we have conducted extensive research on the softgel digital image processing techniques and defect recognition technology.In addition,we has completely constructed the software of special shape transparent softgel detection system.The main work includes:1.We designed a set of special-shaped transparent softgel defect detection schemes and methods.Aiming at the problem of the transparency of softgel,an optimum illumination was selected after comparing the imaging quality of the different color light sources,which enhanced the contrast between the target and the background.Aiming at the problem of blind area caused by the thickness of the softgel shell,a double camera cross detection method is proposed,which can eliminate the influence of blind area in the visual field without adding additional mechanical structure.Choose cost-effective equipment and trigger scheme to ensure that the image acquisition process can be completed efficiently and stably.2.Based on the analysis of various types of defective images,a complete set of image processing algorithms is designed.This algorithm realizes the function of softgel position matching,accurate segmentation of softgel body and background,defect detection and classification.The encapsulating line can be identified accurately to avoid error detection.3.We analyzed the user requirements based on the software engineering method,and realized the complete softgel detection system software based on the C#language.The software functions include camera control,image debugging,parameter selection,user management,data query and other functions,which making users more easily to operate the machine.The corresponding software testing has also been carried out.Experiments show that the defect detection algorithm proposed in this paper has good robustness and it is suitable for different types of fish-shaped softgel.The flaws in fish oil softgels such as missing pill,excess pill,abnormal pill,bubbles,black spots and rubber defects can be detected successfully.The resolution of detection is smaller than 0.05mm and the omission ratio is less than 0.03%,which meet the requirement of special-shaped transparent softgel defect detection.