Sandpaper Defect Detection
Shahid, Imran Ahmad (2021)
Shahid, Imran Ahmad
2021
All rights reserved. This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi-fe2021060232981
https://urn.fi/URN:NBN:fi-fe2021060232981
Tiivistelmä
Sandpaper is an abrasive material used to smoothen, clean, or cut a surface. Sandpaper is made by gluing grains such as garnet or emery with a coating paper. Sometimes grains are not glued adequately with a coating paper due to uneven sheet of stearate layer or other production errors that cause markings on the surface of sandpaper. These marks reduce the efficiency of sandpaper. The goal of this research study is to find a method to classify faulty sandpapers.
Currently, a line scan vision camera is deployed on the production line for quality control. This camera has an algorithm to classify sandpaper. However, this approach is not accurate enough for the industrial standards. The goal is to apply computer vision approaches that include deep learning and a traditional method for detecting defects in the sandpaper. Moreover, this study compares the accuracy of the line vision camera, deep learning, and traditional method.
Currently, a line scan vision camera is deployed on the production line for quality control. This camera has an algorithm to classify sandpaper. However, this approach is not accurate enough for the industrial standards. The goal is to apply computer vision approaches that include deep learning and a traditional method for detecting defects in the sandpaper. Moreover, this study compares the accuracy of the line vision camera, deep learning, and traditional method.