Smart Detection and Reporting of Potholes via Image-Processing using Raspberry-Pi Microcontroller

One of the causes of local road accidents in developing countries, such as the Philippines, is due to road damages such as potholes. In addition, there is no proper road maintenance in the local roads, and so the checking of pothole is done manually. Hence, in this paper we propose a simple and robu...

Full description

Saved in:
Bibliographic Details
Published in2018 10th International Conference on Knowledge and Smart Technology (KST) pp. 191 - 195
Main Authors Garcillanosa, Mae M., Pacheco, Jian Mikee L., Reyes, Rowie E., San Juan, Junelle Joy P.
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.01.2018
Subjects
Automobiles
Cameras
Cloud storage
Global Positioning System
image-processing
internet
microcomputer
Microcomputers
Roads
smart devices
Testing
Web servers
webserver
Online AccessGet full text
DOI10.1109/KST.2018.8426203

Cover

More Information
Summary:One of the causes of local road accidents in developing countries, such as the Philippines, is due to road damages such as potholes. In addition, there is no proper road maintenance in the local roads, and so the checking of pothole is done manually. Hence, in this paper we propose a simple and robust design of a portable and affordable device that will be suitable for local jeepney (cab) drivers here in the Philippines. A distinguishing feature of this proposal is that it does not need a sophisticated Smartphone to automatically send the reports, and was tested in an actual moving vehicle. Furthermore, the system can be installed in a moving vehicle to automatically detect and report potholes via image-processing of Raspberry-Pi microcontroller. Integration of several image-processing schemes has been used to produce an algorithm using Python Language from the OpenCV library that can detect and report potholes automatically from a moving vehicle. The reported image of the pothole and its location are stored and viewed through the use of the Internet, Dropbox, and web server. The system was tested on a Hyundai Eon city car with maximum speed of 10kph-40kph during daytime. With a rate of about 8 frames per second, images were processed per frame to detect potholes by analyzing its color, depth, and area. Overall, the whole system was successfully implemented using the Raspberry-Pi microcomputer and was able to detect and report potholes from a moving car with 100% reporting success rate.
DOI:10.1109/KST.2018.8426203