OPTIMASI PROSES PENGUKURAN DIMENSI DAN DEFECT UBIN KERAMIK MENGGUNAKAN PENGOLAHAN CITRA DIGITAL DAN FULL FACTORIAL DESIGN

Abstract

Theinspection process of surface quality of ceramic tile could be done by using image processing technique through
the optimization by using Center for Ceramics’s parameteron Indonesian National Standard (SNI) ISO 10545.
This research will analyze from light intensities (level 300lx, 600lx, and 900lx), and camera distances (50cm, 75cm
and 100cm), with three times replication using full factorial design. This research uses Matlab 2009a software
to identify area and defect on dry spots ceramic tile’s surface. The result obtained from this research is there were
significant influencing factors: light intensity, and camera distance, as well as the interaction of these factors
towards the error rate percentage of measuring areaand defect on ceramic tile’s surface. The smallest error rate
value from measuring tile’s surface and diameter of dry spots with light intensity of 300lx and camera distance of
50cm had been obtained the error rate value for each measurement about 0.0675% and 2.30%, with combination
of grayscale value for the error rate measurements of tile’s surface and diameter of dry spots were 0.2989 x 0.1140
x R + G+0.5870 x B. Based on the correlation coefficient value between light intensity, camera distance towards
the error rate of measuring areaand defect on tile’s surface, each of them was obtained correlation coefficient value
of camera distance with error rate had 0.518 and 0.516, which meant a strong correlation. The positive correlation
coefficient value showed a unidirectional relationship of two variables, where the rise of one variable would cause
the rise of another variable and the decline of one variable would cause the decline of another variable.