System Analysis Approach to Visual Object Tracking

Abstract

In the modern-day study of system behaviour and analysis, understanding the response of how the system will behave in a certain condition is considered vital. Thus, keeping object behaviour as the primary objective and object detection as a core instrument this project intends to identify the object and track the motion of the desired object. This paper provides an overview of how a closed-loop control-based system was designed using a camera as visual input. The video signal is analyzed to detect distinct color and shape similarities to a desired object. Based on a high similarity index the Raspberry Pi controller generates a pulse width modulation signal to regulate the angular position of the rotary actuator to align the camera frame of view according to a desired object’s position. Post system design experimentation results and potential realized system faults are expressed within the conclusion.