Automatic Detection of Motorcyclist without Helmet using Haar Cascade Classifier

Roopashree Jagannatha Gavadi, Soumya S Patil


Motorcycle accidents are growing throughout the years in all the countries, as there is difference in social, economical and the transport conditions differs from place to place. Mototrcycle is one of the prominent means of transport used by middle class people. Wearing helmet is the main safety equipment of motorcyclists, which might not be followed by all drivers. Accident of a motorcyclist is serious issue. This project aims at prevention of accidents by automatically identifying the drivers wearing helmet or not. For this, a Haar descriptor for features extraction is used. Based on Haar feature Extraction and Haar Cascade Classier real time images captured by cameras are used. The best result obtained from classification was with an accuracy rate of 0.995, and the best result obtained from helmet detection is with an accuracy rate of 0.96.


Feature Extraction; Haar Feature Extraction; HOG Feature Extraction; Cascade Classifier;

Full Text: