Drone App for detection and location of a target or a missing person

Abstract

Nowadays, it is comparatively costly to use helicopters for aerial searches for lost people in town and outskirts than that of using drones. The major disadvantage of the former solution is it requires a lot of resources at a time, such as man power and operating cost. Thus, drones would be a better solution for doing aerial searches for lost persons. With the help of the computer vision technology which uses by drones, it allows drones to detect and locate the targeted individuals in a more effective and efficient manner so as to reduce the cost of search and rescue. If multiple drones are used for searching lost persons at the same time, it can even improve the effectiveness and efficiency of rescues. In this project, the study focuses on implementing the basic functions that make the drone possible to do aerial search, that is, to detect and locate the lost ones by using computer vision technology. An app named "DJI Detect" is designed and implemented for searching by making use of the drone. There are mainly two functions provided by the app. They are: 1) Identifying a targeted person based on the colour of the target's clothes and pants, and 2) Reporting the target's current location on map. After doing several tests with the app, the results are matched with the expected outcomes of the project. The app "DJI Detect" can detect a targeted person successfully when the drone is flying in the air. Also, it can report the location of the target accurately. In order to enhance the effectiveness and efficiency on drones’ search and real rescue service, one of my suggestions is drone developers can try to work on multiple drones’ cooperation.