Utilization of Complex Techniques for the Extraction of Human Body Parameters from Uncalibrated Image

The principal goal of this paper is to propose a technique that will allow extraction of human body parameters from uncalibrated image, having some known reference height. Analysis focuses on the algorithms for detection of objects from the image. The designed approaches were experimentally tested in a form of a software solution that utilizes vanishing line height extraction approach. The system allows to measure all human body skeleton model parameters from the image with known height of reference object. System utilizes recently proposed Human Body Description Format (HBDF) to store the measured results.