View Invariant Motorcycle Detection for Helmet Wear Analysis in Intelligent Traffic Surveillance
An important issue for intelligent traffic surveillance is automatic vehicle classification in traffic scene videos, which has great prospective for all kinds of security applications. Due to the number of vehicles in operation surpassed, occurrence of ac
- PDF / 386,745 Bytes
- 11 Pages / 439.37 x 666.142 pts Page_size
- 14 Downloads / 156 Views
Abstract An important issue for intelligent traffic surveillance is automatic vehicle classification in traffic scene videos, which has great prospective for all kinds of security applications. Due to the number of vehicles in operation surpassed, occurrence of accidents is increasing. Hence, the vehicle classification is an important building block of surveillance systems that significantly impacts reliability of its applications. It helps in classifying the motorcycles that uses public transportation. This has been identified as an important task to conduct surveys on estimation of people wearing helmets, accident with and without helmet and vehicle tracking. The inability of police power in many countries to enforce helmet laws results in reduced usage of motorcycle helmets which becomes the reason for head injuries in case of accidents. This paper comes up with a system with view invariant using Histogram of Oriented Gradients which automatically detects motorcycle riders and determines whether they are wearing helmets or not.
⋅
⋅
Keywords Background subtraction Histogram of Oriented Gradients (HOG) Center-Symmetric Local Binary Pattern (CS-LBP) K-Nearest Neighbor (KNN)
⋅
M. Ashvini (✉) ⋅ G. Revathi ⋅ B. Yogameena Department of ECE, Thiagarajar College of Engineering, Madurai, India e-mail: [email protected] G. Revathi e-mail: [email protected] B. Yogameena e-mail: [email protected] S. Saravanaperumaal Department of Mechanical, Thiagarajar College of Engineering, Madurai, India e-mail: [email protected] © Springer Science+Business Media Singapore 2017 B. Raman et al. (eds.), Proceedings of International Conference on Computer Vision and Image Processing, Advances in Intelligent Systems and Computing 460, DOI 10.1007/978-981-10-2107-7_16
175
176
M. Ashvini et al.
1 Introduction Recently, detecting and classifying moving objects from video sequences has become active research topics. They are used in various circumstances nowadays. Segmenting and classifying four moving objects such as bicycles, motorcycles, pedestrians and cars with view invariant in a video sequence is a challenging task. The object can be detected both in motion as well as in rest position depending on the application. Despite of its significance, classification of objects in wide scenario surveillance videos is challenging because of the following reasons. As the capability of conventional surveillance cameras [1] is limited, Region of Interest (ROI) in videos may be of low resolution. As a result, the information supplied by these regions is very limited. Also, the intra class variation for each category is very huge. Objects have diverse appearances and they may vary significantly because of lighting, different view angles and environments. The potential for object classification in real time application is great and so its performance has to be improved. However, the above mentioned issues reduce the accurate working of object classification algorithms. Helmets are essential for motorcyclists’ security from deadly accidents. The inabi
Data Loading...
