Ribbed Smoked Sheet Grading System (RSSGS)

Abstract

Currently in the rubber factories employ many experts or scientists to grade the Ribbed Smoked Sheet (RSS). They grade the RSS by using their eyes and experience. The objective of this research is to build a computer system that can help rubber experts to grade the RSS. This system is called “Ribbed Smoked Sheet Grading System (RSSGS)”. The system consists of 4 main components, which are 1) Image Acquisition, 2) Image Preprocessing, 3) RSS Grading, and 4) Display of Result. In the image acquisition component, we use a digital camera to take an RSS image in a controlled environment box. In the image preprocessing component, we apply several image processing methods to prepare a suitable RSS image for a grading process. In the RSS grading component, we apply the k-Mean Clustering and the Euclidean Distance method to classify the RSS into five grades, which are RSS1, RSS2, RSS3 RSS4 and RSS5. In the Result Display component, we create a graphic user interface (GUI) for displaying results of the RSS grading. We test the system by using 398 RSS images for a training dataset and another 322 RSS images for an un-training dataset. The precision rates of our RSSGS are 80.90 percent for an untraining dataset. The average access time for the RSSGS is around 10.88 seconds per RSS image.