Tablets and Capsules Blister Pack Inspection

Kang, Yan Sheng (2023) Tablets and Capsules Blister Pack Inspection. Final Year Project (Other), Tunku Abdul Rahman University College.

Text Report-KANG YAN SHENG.pdf Restricted to Registered users only Download (36MB)

Abstract

Tablets and capsules are the most widely used pharmaceutical dosage form. During the packaging of tablets and capsules into blister packs, there might be some situations where the tablets and capsules are broken or missing in the blister pack. This project is made to detect missing and broken tablets and capsules in the blister pack. The Raspberry Pi hardware will be used in this project in conjunction with OpenCV, the open-source computer vision library. Furthermore, a conveyor belt will be used to move the package to a fixed location for inspection. K-means clustering is used to segment the image, and the segmented image is then used to find the contours of tablets and capsules. A JSON file consisting of the data of the samples is used for comparing the features of the inspected sample. The good tablets and capsules will be indicated with a green bounding box; defective tablets and capsules will be indicated with an orange bounding box; the missing tablets and capsules will be indicated with a red bounding box. By implementing this system in the manufacturing line, productivity and efficiency can be increased as this machine has automated the inspection process. As a result, defective blister packs will not be distributed to consumers