Harijs Grinbergs, Artis Mednis, and Modris Greitans. Real-time object tracking in 3D space using mobile platform with passive stereo vision system. N. Tagoug (Ed.): Proceedings of World Congress on Multimedia and Computer Science (WCMCS 2013), pp. 60-68, 2013. Association of Computer Electronics and Electrical Engineers, 2013.
Tracking of objects in 3D space performed in real-time is a significant task in different domains from autonomous robots to smart vehicles. Traditional methods for this task include the usage of specific data acquisition equipment as radars, lasers etc. The expansive development of computer technologies enables to solve this task using general purpose data acquisition equipment as stereo cameras and smart image processing for object tracking. The paper is describing a passive stereo vision based mobile system for real-time object tracking in 3D space. Selected image processing algorithm is discussed and its evaluation on real world data presented with tracking time per target as short as 0.1-0.2ms in urban environment traffic scenarios. The initial set of algorithm parameters and constraints is determined as well as some recommendations for their application is given.