Development Of Object Tracking Architecture Based On Optical Flow Vector For Autonomous Motion Detection System

Yusoff, Naszeri (2016) Development Of Object Tracking Architecture Based On Optical Flow Vector For Autonomous Motion Detection System. Project Report. UTeM, Melaka, Malaysia . (Submitted)

Text (24 Pages) Development Of Object Tracking Architecture Based On Optical Flow Vector For Autonomous Motion Detection System.pdf - Submitted Version Download (378kB)

Abstract

Autonomous motion detection system is important in many developed application. The applications of object tracking can be utilize in various categories for a numerous important reason such as humans, vehicles random and other random objects movement. The motion captured from the video stream will be analyze and the direction and speed of the dynamic object have to be obtained. Hence, the suitable method of calculating the optical flow need to be chooses wisely. This research is to identify moving object and dominant moving direction for each of motion vector and analyze the data within execution time. Then, to process a video frame by frame as they appear in video stream and display in GUI. The system need to capture object motion from the video stream and display it using GUI which in this research Myrobotlab software will be used as the system interface. By implementing optical flow technique, the system has to model and provide data collection of motion existed. The pattern modeling motion of the object has to be analyzing in order to determine direction and speed of dynamic object motion. OpenCV library will be integrated in Myrobotlab as well as Arduino which act as a microcontroller between Webcam camera and the GUI. The algorithm which is Lucas Kanade is used as part of the system. The results provide the motion direction and speed of the camera target by the system. The technique is successful to detect object and provide the parameter needed to display the processed motion existed. In general, the developed system can achieve the objective to track an object and being used in related vision technology field.

Actions (login required)

View Item