ESTIMATION HUMAN HEIGHT FROM VIDEO CAMERA

• Main Author: HARLEY, EDWIN@JOHNNY
• Format: Final Year Project Report
• Language: English; English
• Published: Universiti Malaysia Sarawak, (UNIMAS) 2019
• Subjects: TK Electrical engineering. Electronics Nuclear engineering
• Online Access: http://ir.unimas.my/id/eprint/34328/1/Harley.pdf; http://ir.unimas.my/id/eprint/34328/4/Harley%20full.pdf; http://ir.unimas.my/id/eprint/34328/

This paper presents about estimation of human height with video camera. There have been many methods that developed to measure accuracy of human height algorithm with image processing techniques in computer such as non-contact measurement, RGB-Camera, calibrated camera and etc but some of this technique have weakness of the can be detected such as errors. However, uncalibrated camera technique is chosen in my research because it can measure human height automatically, reliable, and can be proved when capture and record the image of subject with using camera of smartphone in multiple images. Therefore, uncalibrated camera technique is implemented in my research project for measurement of human height with video camera from any smartphone. There are three steps for analysis human height algorithm measurements such as vanishing point, line extraction, and image processing technique that involved in my research of estimation of human height. Moreover, MATLAB R2018a is software that be used for process of extraction images because it would execute the value of pixel for each subject and read of all subject’s image. Process of extraction images is important because it can subtract image of subjects and image of background such as extract head of human to foot for ensure the subjects can be recognized as a human and non-human. Vanishing point is needed to calculate the height so that it is accurate and reliable. Finally, uncalibrated camera method is proposed based on image background and image of walker.