PSO-particle filter-based length estimation in human tracking

Abstract

Today, high quality results are expected of security and surveillance systems. These systems must not only track the real sizes and motions of humans, but also, in some situations, they must track features such as width and length. Few methods have been proposed for height surveying. Some studies show that an infrared ray technique can survey the height of a human, but the equipment required is complicated. The objective of this paper is to build a mathematical model and method for human tracking to solve the problem of length surveying. This human tracking method can mark humans’ and objects’ size in a picture so that, if we put this picture in a frame of axes, we can calculate the human’s/object’s height or other lengths. To obtain more accurate results for height surveying, we need a tracking method that can show more exact results. Combining tracking/detecting methods with a pso-particle filter will show a result with great accuracy in human tracking.