Automatic Pothole Detection and Navigation System Using Smartphone

Abstract

With widespread adoption of smartphones, a number of human’s robust issues square measure being tackled with mobile applications. one in all several such issues is road traffic accidents, and engorged traffic in metropolitan cities. in keeping with the National Crime Records Bureau (NCRB) road accidents accounted for one, 54,732 fatalities in Asian nation within the year 2019 within which fifty % accidents were because of dangerous road condition and with growing range of car users, traffic is growing day by day. Technology, a lot of specifically, itinerant technology has evolved to modify miniature devices the aptitude of containing powerful sensors. The functionalities of those sensors, like accelerometers, gift in smartphones is what this study exploits to develop a system capable of mechanically police investigation potholes in time period and observance road traffic conditions. Machine learning techniques; Support vector machines supported K-means bunch, square measure applied to the info obtained from such sensors to estimate road/traffic conditions. Previous add this space puts the incumbrance on the user, and pays very little attention to giving incentives for the tedious, and mundane task of cataloging potholes, or the other road anomalies. As such, this study goes on the far side merely police investigation or estimating road/traffic conditions and derives utility for the user by creating use of the info collected to modify hindrance of potholes whereas driving, and visualizing roads traffic which might inform selections on alternate routes. The developed system is evaluated exploitation knowledge obtained from the crawdad.org information and a check drive on Ilorin roads shows promising results.