Potholes, debris, sunken manhole covers and others are common street safety hazards which drivers experience daily as they bump into them unexpectedly while driving. The International Roughness Index (IRI) is the most prevalent metric that is used to evaluate pavement roughness in transportation agencies. This study primarily basis on the classification of IRI values that are collected by using an android smartphone application using Artificial Neural Networks (ANN) for the detection and analysis of diverse predefined street safety hazards and classification as disorder or normal area. The neural network designed in this study is the back-propagation type and trained by using the Gradient Descent (GD) training algorithm. Before the classification process, the IRI values are pre-processed for extracting some features from them. The conventional and most effective features are extracted and normalized. The neural network is trained with the normalized feature set by using supervised learning method. Simulation results show that the designed network can successfully classify the street conditions by using IRI values with a satisfying success rate. The performance of the designed network is compared to a similar research work that is previously presented in the literature and supported with a detailed performance comparison demonstration. The performance of the designed network is compared to a similar research work that is previously presented in the literature and supported with a detailed performance comparison demonstration.
Click to View
Click to View
Click to View
Click to View
Click to View
Click to View
Click to View