Asian Journal of Engineering, Sciences & Technology

Human Detection and Counting in Crowded Scene

Asian Journal of Engineering, Sciences and Technology - Volume 4, Issue 1 2014
By Pirah Noor Soomro, Ufra Memon, Sheeraz Memon
Keywords: head detection, people counting Ada BoostDetection Cascade.

Crowd surveillance is an active topic ofresearch nowadays. Increased law and order situations have greatly alarmed the security organizations to improve security services in all the fields especially when security of crowd is brought under consideration. Crowd surveillance is different. In crowd there is congestion, noise and large number of people which are distributed in random manner which cause much difficulty for security officers. Therefore now crowds are scanned by cameras and are processed in computer for surveillance and for getting information of the scene. Many approaches have been proposed in this context. Keeping in view the complexity of scene in the domain of computer vision we come across the big problem of occlusion. The ultimate solution needed is the technique which overcomes the problem of occlusion. In this paper we have proposed a technique, which provide good hand in surveillance. Our approach has two folds, we detect the heads of the people in crowd and then we count the people present in the crowd. Detection of heads is done by the famous Viola and Jones Method and counting is done based on the output of detection.

Share this paper


Want to publish in Asian Journal of Engineering, Sciences and Technology?
Send us your paper for review
Authors
Research Papers
Citations