Machine Vision Inspection System for Detecting Coating Defects in Metal Lids

Experts forecast that globalization and open economy will be the characteristic of the 21 century. Industry in developing countries must be well prepared and competitive in order to have a vital role. These countries will not be able to access global markets unless they procure the needed technology to manufacture high quality, up to the standard products. Automated visual inspection is one of these emerging technologies. Higher production speeds require higher inspection speeds which can be implemented best by replacing human visual inspection procedures by machine vision inspection systems. This paper presents an online machine vision system design that is capable of detecting defects in rubber coating of metal lids during manufacturing. The system is simple and low cost consisting of a CCD camera mounted on a conveyer belt and connected to a PC through FireWire (IEEE 1394). Image acquisition, analysis, and inspection are implemented using National Instrument’s (NI) Vision Builder tool. The inspection algorithm is realized through a four stage process: starting with acquisition to enhancement, thresholding and finally object matching.