<![CDATA[One factors of the road repair process that takes a long time caused by the process of recording the condition of road damage that is still done manually by human labor. Along with the development of technology, many research related to road damage detection system using digital image processing. The purpose of this research is to build a retrieval system of asphalt pavement image based on damage level. The process begins with pre-processing to get a segmented hole area. Furthermore, utilizing feature extraction of Gray Level Co-occurrence Matrix (GLCM) texture. Features used in this research are as many as 52 features derived from 13 features with angles 0o, 45o, 90o and 135o. Of the 52 features performed feature selection using Wrapper and CFS (Correlation Based Feature Selection) methods. Based on the results of the tests that have been done we get the image of 117 holes that successfully segmented successfully on the diameter of 101x101, = 75 and =75. Use of the Wrapper feature selection method gives higher average accuracy and MAP (Mean Average Precision) results than using the CFS feature selection method or not using feature selection. Accuracy and MAP resulting from Wrapper method with d = 1 respectively that is equal to 55.61% and 0.710.]]>
Copyrights © 2018
