CLASSIFICATION OF ROAD DAMAGE USING SUPERVISED LEARNING TO ASSIST VISUAL ASSESSMENT OF ROAD DAMAGE

• Main Author: Nurrosyid Al Haqi, Novindra
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/68965
• Institution: Institut Teknologi Bandung

Roads are an important aspect in improving the welfare and productivity of the people that affect the country's economic growth. Therefore, monitoring of road conditions must be carried out. The Department of Highways and Spatial Planning of West Java Province in collaboration with the Bandung Institute of Technology has developed the SKPJ (Road Pavement Condition Survey) application which makes it easier to monitor road conditions. However, some features of this application are still carried out semi-automatically, one of which is the road damage classification feature. This causes long working time. Classification of road damage itself needs a visual assessment by humans. Therefore, a solution is needed in the form of a machine learning-based road damage classification system that can help users visually assess road damage. The system uses the supervised learning method to build a classification model. The system still requires a visual assessment by the user to validate the system classification results. The system that has been built is considered to be able to help classify road damage by assisting the user's visual assessment, although the classification results by the system have not been able to provide a correct classification of road damage.