< Projeler Sayfası

Proje Adı
Çoklu Kamera Takip Sistemi

Şirket Adı
Karel

Şirket Mentoru
Dr. Nail Çadallı

Akademik Mentor
Prof. Enis Çetin

Asistan
Onur Külçe

Takım Üyeleri
Yavuz Selim Mutlu, Zülkarneyn Şişman, Fatih Eroğlu, Kazım Görgülü

Summary
People use GPS technology to determine 3D position of an object in open areas, but in an enclosed space GPS technology cannot be used. At this point, there occurs a need for determining the 3D position of an object in a closed environment. In this project, under the leadership of Karel, a multi-view video camera system that can determine the three dimensional (3D) positions of a number of (reflective) moving objects within an enclosed space in real time by processing the video image sequences from various different view angles is developed. Methodology of this project can be divided into two parts which are hardware and software. 4 high definition cameras with 30 frame per second (fps) processing speed is used as hardware. At the installation part of this system, those 4 cameras is placed on the walls of a rectangular room at any position and view angle. The software part of this project can be divided into two parts as well which are object detection and object tracking. Object detection is designed as it can get 3D positions of the moving objects correctly and it is based on color detection. Object tracker is capable of tracking multiple objects simultaneously and it is also smart in terms of handling the occlusions. Finally, detected 3D locations of the moving targets will be shown to the user by using graphical user interface. This system can determine 3D positions of multiple objects with at most 1% error and can be used in several areas like military, robotic systems such as tracking unmanned vehicles in a closed environment.