| Title |
Road visualization for smart city: Solution review with road quality qualification |
| ID_Doc |
40804 |
| Authors |
Leduc, E; Assaf, G |
| Title |
Road visualization for smart city: Solution review with road quality qualification |
| Year |
2020 |
| Published |
|
| DOI |
10.1016/j.iot.2020.100305 |
| Abstract |
Pavement management consists of a series of maintenance and rehabilitation activities over the life of a road. Maintenance involves different methods and technologies that extend life by slowing the rate of deterioration. Currently, the structural conditions of the road network are assessed in the traditional way, that is, it is done manually by technicians using specialized tools. Thus, to measure deflection, for example, we use a deflectometer or a Benkelman beam. If you want to assess the deterioration of the road surface or count the vehicles passing on a road, then technicians must make a visual analysis of the road. In the same vein, most of the available research evaluates the results of data extraction and does so mainly visually. So, when a frame of reference is available, an adjustment is made to modify the database. This is where IoT (Internet of Things) comes in place by creating a simple automation process to standardize the data and transform them in information to create a database and measure the quantity and quality of the deformation with a Neural Network. The data set has been categorized manually into different types of deterioration, allowing it to be measured quantitatively and qualitatively. The sample comes from images collected by a GoPro camera, on a 60-degree angle, at 15 FPS. 4 different types of deformation were present in the data set; our model predicted with a precision ranging between 50% and 90% the different types of deformation. The validation of the results of the model shows 71% of true positive. This process automation provides standardized information for a road qualification and quantification system. Crown Copyright (C) 2020 Published by Elsevier B.V. All rights reserved. |
| Author Keywords |
AutoML; Data acquisition; International roughness index; Machine learning; Road quality; Sensors |
| Index Keywords |
Index Keywords |
| Document Type |
Other |
| Open Access |
Open Access |
| Source |
Science Citation Index Expanded (SCI-EXPANDED) |
| EID |
WOS:000695695600021 |
| WoS Category |
Computer Science, Information Systems; Engineering, Electrical & Electronic; Telecommunications |
| Research Area |
Computer Science; Engineering; Telecommunications |
| PDF |
|