Abstract: Now a days, the number of accidents increased due to the increase of number of vehicles. The driver must be known about the road condition for his safety. There are some expensive methods to install a dedicated hardware on the vehicle. So we have designed a method based on smart phone using Accelerometer and GPS sensors to identify the road condition. The system that is designed is called Bumps Detection System. The potholes are detected by using the accelerometer and the location of potholes are plotted on the map by using the GPS sensor. So we implemented an Android application by extracting the API of Google Map to locate the potholes in the roads by using a smart phone device held by the driver. The condition of the road and the number of potholes will be informed to the driver in advance in the map by a warning mark. While designing the system, some threshold values that are experimentally derived. A machine learning approach is used to find those threshold values. By using these information, the system is trained. To build the system on the trained data, K-means algorithm is applied. For better prediction, the Random Forest Classifier is applied on the test data. In this application we also include voice based command in both English as well as in our regional language (Malayalam) to notify the driver about the condition of the road.
Keywords: Accelerometer, Google Map, GPS, K-means Clustering, Random Forest Classifier
|
DOI:
10.17148/IJARCCE.2020.9119
