Sebuah Sistem Pencitraan Optik berbasis pada Proyeksi Pola Titik-titik Acak dengan Metode Digital Image Correlation (DIC) untuk Menentukan Profil Kekasaran Permukaan suatu Bahan

Abstrak

Reliable methods (techniques) are needed to determine surface roughness accurately. Contact techniques are prevalent but can damage materials and are subject to much noise. Meanwhile, non-contact techniques (usually optical) still face setup complexity and inaccuracy. This paper designs a simple optical imaging system based on the projection of random dot patterns to determine the surface roughness of a material. It is technically simple as it only requires a Liquids Crystal Display (LCD) projector device to highlight the random dot pattern onto the surface of the test material and then a digital camera to capture it. The camera captures a reference image (just the random dots pattern) and a distortion image when the random dots pattern highlights the test material. Both images are digitally processed using the Digital Image Correlation (DIC) algorithm to obtain a correlation value proportional to the change in the pattern of random dots due to the presence of the test material surface. This change manifests as a pixel displacement field image estimated to be proportional to the surface roughness profile of the material. Based on this profile, a surface roughness index can be estimated. The validation test results show that the design can determine the roughness index ( of a randomly sprinkled plane surface with different diameters of cork balls (selected). The graph result shows that index  tends to increase linearly with the increase of cork ball diameter with a correlation value of = 0,97. The larger the diameter of the cork ball, the larger the surface roughness index. This result is rational and valid. The design is also verified to determine the roughness index of the surface of commercial sandpaper with various grit sizes. The larger the grit of the sandpaper, the smoother the sandpaper. The verification results show that the design can demonstrate that the larger the grit, the smaller the roughness index of the sandpaper. Thus, the developed optical imaging design is valid and has the potential to be simple, cheap, easy to operate, and fast in obtaining information about the surface roughness of a material.

Keywords: imaging, optical, projection, random-dots, surface-roughness.