Development on Autonomous Object Tracker Robot using Raspberry Pi

Abstract

Surveillance systems is widely used to help human in preventing crimes, ease humans in surveillance work, enabling remote monitoring and improving productivity. Previous research paper has developed a surveillance system using RCNN which enhancing the implementation of detecting object and motion in that appears in the camera. There are also several papers that stated the implementation of sensors in detecting any appearance of human and objects which are necessary in the security system in various industries such as military and agriculture fields. This study is focusing on developing a robot surveillance system that are able to be detecting motion and object for a small area such as home and office as it is utilizing Wi-Fi connection for the robot and the controller. Therefore, the robot development would be implementing Arduino and Raspberry Pi for the main hardware components while the algorithm used for software in code development are the combination of OpenCV and the COCO Dataset to implementing the detection of the object and motion for the camera. Robot body are developed from scratch to suits the compatibility of the other components and boards. The findings that have been found is that the lower specification on the Raspberry Pi board has causing the camera frame per second is decreased, which resulting in live video lags and unsynchronized video output.