Vision-based liquid level detection in amber glass bottles using OpenCV

• Main Authors: Felipe, Mikhael Anthony A., Olegario, Tanya V., Bugtai, Nilo T., Baldovino, Renann G.
• Format: text
• Published: Animo Repository 2019
• Subjects: Image processing; Bottling; OpenCV (Computer program language); Manufacturing
• Online Access: https://animorepository.dlsu.edu.ph/faculty_research/2452; https://animorepository.dlsu.edu.ph/context/faculty_research/article/3451/type/native/viewcontent
• Institution: De La Salle University

All manufacturing processes, from raw material processing to electronics fabrication, require quality control to ensure consistency across all products of same specifications. It helps establish the reputation of a certain brand, gain customer loyalty, and maximize profit. In the bottle filling industry, it is important to ensure that the amount of product in each bottle is consistent with the packaging label. Less than that, the company may lose customers and face legal consequences; more and the company loses profit by giving more than what was marketed. This study makes use of a vision-based technique in detecting the liquid level in amber glass bottles. The proposed system has applied Python and OpenCV for the pre-processing and image processing. Moreover, the study was successful in detecting and classifying filled bottles into three categories: under-fill, within target and over-fill. © 2019 IEEE.