Title :
Roll-to-roll web handling for membrane inspection
Author :
Tan, Henry ; Tay, Arthur ; Ng Bing Qiang ; Chong Wee Keat ; Kiew Choon Meng ; Chow Siew Loong
Author_Institution :
Dept. of Electr. & Comput. Eng., Nat. Univ. of Singapore, Singapore, Singapore
Abstract :
This paper describes our ongoing development of a roll-to-roll (R2R) web handling system for thin and flexible membrane inspection using machine vision. Leakproof membrane is essential in many industries, e.g. food packaging and medical applications. Our primary objective is to achieve full quality control check for through-hole defects in membrane, and mark them if any. To achieve 100% inspection of the 400 mm wide and 1 km long web, a laterally movable camera stage is coordinated with web handling subsystem that uses basic trapezoidal motion control for repositioning the web so as to overcome the small area of inspection of the two line scan cameras mounted on stage, each with 80 mm field of view. As there is no open-source publication on web handling for R2R membrane inspection, we hope our methodology and findings through this work would be useful to other related research. Besides detailing the system design and setup, this paper also discusses the challenges faced, highlights the system performance, its shortcomings and areas of improvement.
Keywords :
computer vision; control engineering computing; inspection; membranes; motion control; production engineering computing; quality control; R2R membrane inspection; R2R web handling system; camera; flexible membrane inspection; food packaging; leakproof membrane; machine vision; medical applications; membrane inspection; roll-to-roll web handling; system performance; thin membrane inspection; trapezoidal motion control; web handling subsystem; Acceleration; Cameras; Coatings; Inspection; Manufacturing; Motion control; System analysis and design; Roll-to-roll (R2R); high resolution membrane inspection; precise motion control; speed control; web handling; web inspection;
Conference_Titel :
Control Automation Robotics & Vision (ICARCV), 2014 13th International Conference on
DOI :
10.1109/ICARCV.2014.7064359