Semi-Autonomous Human Detection, Tracking, and Following Robot in a Smart Building
Parth Mannan, Keerthi Priya Pullela, V. Berlin Hency, O.V. Gnana Swathika · 2021
Surveillance and monitoring in a smart building can usually become a difficult task to carry out. Multiple factors, including weather conditions, venal nature of humans, and risky conditions – such as the presence of land mines – make effective monitoring a challenge. To address these challenges, mobile robots can offer an effective solution. The proposed work aims to address these challenges by building a semi-autonomous system that can detect human presence effectively and also track the human until the identity is verified. The proposed work uses the well-known method “histogram of oriented gradients” (HOG) for human detection. For tracking and following, the algorithm built has been based on the multiple instance learning (MIL) and Lucas-Kanade (LK) optical flow tracking method. A password protection module has been included in the system in order to reset the robot once the trespasser has been caught by the authorities or a friendly person was detected as a possible trespasser. The system has been implemented and tested on the Raspberry Pi board running the OpenCV library.