Abstract—Pedestrian detection is a challenging problem. There are various researches to detect people in digital images. Besides, to recognize children from adults in digital platforms is helpful for applications which are created to people’s good account. For instance, if Closed Circuit Television (CCTV) which is located on traffic lamb detects child who is walking through pedestrian way, system could make some adjustment in a good way. Aim of this article is to detect children and adults separately in digital images. While we employed our method, we applied Haar Cascades which is widely used technique in object detection. At first, we detected head and full body of pedestrians, then we used relative measurements. So we did proportioning head size to body size of pedestrians. By this technique, we tried to discriminate children and adults. Our results were not %100 accurate but it gave a clue to improve ourselves. It seems that it’s improvable for now.

Index Terms—Absolute measurement, adult and child classification, Haar-like feature, pedestrian detection, relative measurements.

The authors are with the Department of Electronics Engineering, Kyungsung University, Busan, Korea (e-mails: faruk21@ks.ac.kr, jsipark@ks.ac.kr, jsong@ks.ac.kr, bwyoon@ks.ac.kr).

Cite:Omer F. Ince, J. S. Park, J. Song, and B. W. Yoon, "Child and Adult Classification Using Ratio of Head and Body Heights in Images," International Journal of Computer and Communication Engineering vol. 3, no. 2, pp. 120-122, 2014.