Human Distance Estimation Using Quadcopter For Surveillance Purpose

• Main Authors: Shukor, Ahmad Zaki, Khaw, Huai Jian, Ali @ Ibrahim, Fariz, Mohd Aras, Mohd Shahrieel, Jamaluddin, Muhammad Herman
• Format: Article
• Language: English
• Published: Blue Eyes Intelligence Engineering & Sciences Publication 2019
• Online Access: http://eprints.utem.edu.my/id/eprint/24252/2/L109210812S219.PDF; http://eprints.utem.edu.my/id/eprint/24252/; http://www.ijitee.org/wp-content/uploads/papers/v8i12S2/L109210812S219.pdf
• Institution: Universiti Teknikal Malaysia Melaka

Nowadays, quadcopters are commonly used. Quadcopters are unmanned aerial vehicles with four propellers to provide lift to fly and hover above ground. Quadcopter nowadays is a very common commercial item in everyday life. Some quadcopters are designed to do 3D or 2D mapping of a certain area or to take videos or just for entertainment purposes. Quadcopter is a very versatile item and is able to change into anything for example a quadcopter can also be used for security purposes to decrease the crime rate of our country. The objective of this study is to design and develop a quadcopter with image processing system to have the ability to measure the distance of a human from the drone itself. The quadcopter is designed to be small in size and have a mini computer like Raspberry Pi on top of it to compute the algorithm to calculate the distance of the human by using image processing technique through the camera which is setup on the drone. Human detecting algorithm YOLO and software Open CV is chosen to detect human and calculate the distance from the quadcopter. The results show that the system is quite limited by the capabilities of the hardware. The system shows an accuracy of more than 90 percent when the human is standing within a certain range. Both the accuracy of the distance sensing and human recognizing system is affected by the limitation of the hardware.