Abstract In this paper, we propose an approach to automatically identify the bike riders who are not wearing helmets. This approach takes video feed from the surveillance camera deployed at roads and applies a background subtraction technique to identify moving vehicles. From the foreground blobs, different features are extracted to identify motorcycles among the other vehicles. From the motorcycle objects, the head region of the blob is considered to extract helmet-related features. For the performance and accuracy improvement, we apply principal component analysis (PCA) on the derived features. To detect helmet from the motorcycle object, we apply different machine learning techniques on the selected features and perform the feasibility analysis.
An Enhanced Approach for Detecting Helmet on Motorcyclists Using Image Processing and Machine Learning Techniques
2018-07-06
11 pages
Article/Chapter (Book)
Electronic Resource
English
An Efficient Approach for Detecting Helmets on Motorcyclists Using Machine Learning Techniques
Springer Verlag | 2018
|Are helmet laws protecting young motorcyclists?
Online Contents | 2007
|Motorcyclists helmet noise - measurement and attenuation
Automotive engineering | 1994
|Head injuries of motorcyclists with and without helmet protection
Automotive engineering | 1985
|Effect of the helmet act for motorcyclists in Thailand
Elsevier | 2001
|