Color and Shape Based Automatic Detection of Pedestrians in Surveillance Videos

Volume 5 Issue 1 January - March 2018

Research Paper

Color and Shape Based Automatic Detection of Pedestrians in Surveillance Videos

Harihara Santosh Dadi*, Gopala Krishna Mohan Pillutla **, Madhavi Latha Makkena ***

* Associate Professor, Department of Electronics and Communication Engineering, Aditya Institute of Technology and Management, Hyderabad, Andhra Pradesh, India.

** Professor, Department of Electronics and Communication Engineering, Institute of Aeronautical College of Engineering, Hyderabad, Andhra Pradesh, India.

*** Professor, Department of Electronics and Communication Engineering, JNT University, Telangana, India.

Dadi, H.S., Pillutla, G.K.M. and Makkena, M.L. (2018). Color and Shape Based Automatic Detection of Pedestrians in Surveillance Videos. i-manager’s Journal on Image Processing, 5(1), 1-11. https://doi.org/10.26634/jip.5.1.14612

Abstract

Detection of human is the first and foremost step in any human tracking system. Color and shape based automated human detection algorithm in surveillance videos have been proposed in this paper. Every frame is first divided into R, G, and B frames. The background images for these three basic colors are formed, and three binary images for three colors are formed for every frame. One binary image overlaps over the other to find the common rectangles. Based on the aspect ratio of the rectangle, humans are detected. The proposed algorithm uses both color and the shape aspect in finding the humans. The performance results show that the proposed algorithm has good metrics.