Design of Marker Displacement Measurement Systems with Digital Image Correlation from Monocular Camera at Laboratory Scale

• Main Author: Gerry Akbar, Ramadhan
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/43878
• Institution: Institut Teknologi Bandung

Digital cameras can be used to measure the distance of displacement of all types of objects with the rules of digital image correlation, ie comparing images before and after being moved. This study uses marker to simplify the selection of point for being observed. The limits of the ability of such a measurement system depend on the combination of the camera, lens, size of the marker, and the distance between the object and the camera. Thus, this method is considered practical and relatively inexpensive because the measurement range depends on how the object is projected on the camera plane. The maximum target of measurement error in this study is 1 mm. Measurement data from digital images is obtained by photogrammetry or image data processing into spatial equations. The image contains a marker attached to the bridge deflection observation point. Before that, the camera needs to be calibrated and the markers need to be defined in the program. The results obtained are euclidean distance (three dimensions) displacement the center of the marker with a unit of length. The measurement results are also the optimization writer by using an edge-preserving filter, which is a low pass filter but still maintaining the outline, and a digital stabilizer of digital image. The achievement of this measurement system can be known by knowing the error in the validation results between the calculation results and the actual conditions. The error can also be obtained from the error distribution to determine the level of accuracy, precision, and accuracy. The results of this study show that from 90 data collection and average accuracy of 0,29 mm/pixel, measurement errors are less than 2 mm and can be optimized to less than 1 mm when using image stabilizers and edge-preserving filters. The probability of an error within the accuracy of the measurement system increases significantly from 0,4 to 0,7.