Identifikasi Label Kode Pipa pada Sistem Konveyor untuk Pipe Handling Berbasis Template Matching

Abstract

Nowadays, technological advancements are growing need for computer programs that might benefit
industrial purposes. Image recognition is crucial in technological advancements spanning from medical demands
to security and other industries. One of them is the necessity in the oil and gas industry for a computer
application that can recognize pipe code labels, particularly in pipe handling. Because of the requirement for
production assistance, to recognize the pipe code label image, an NI Vision control system was built. There are
numerous stages of label identification in this study, including taking photographs with a webcam camera
(capture), cropping, grayscale, and saving data. The system can provide an accuracy of 88% based on test results
on 25 samples of ¾ inch PVC pipe code labels with a camera distance to the pipe code label field of 85 cm, 493
color weights, and an optimal score level of 75%.