Unmanned Aerial Vehicles (UAV) commonly called as “Drones” are pilotless aircrafts which can be used in areas where manned aircrafts are too risky or difficult to operate. They are basically flying Robots which can be manually controlled or made autonomous with the help of software-controlled flight plans. They fly with the help of inputs received by the Embedded Processor from On Board sensors and Global Positioning System (GPS). Usage of Drones are on increasing demand with each year passing because of the wide variety of practical and innovative uses and applications offered by them. There are certain challenges faced by drones irrespective of the application in which they are being used. One of the main challenges faced by them is to detect and avoid objects which come in their path during the navigation. In this paper, we propose a deep learning based frontal object detection using pre-trained neural networks. The image frames obtained using the front facing monocular camera of the drone are processed and fed into the deep learning network for object detection. An overview and comparison of three deep learning algorithms in terms of their accuracy and speed is also considered.
Drone Object Detection Using Deep Learning Algorithms
02.09.2021
793677 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Drone-Based Weed Detection Architectures Using Deep Learning Algorithms and Real-Time Analytics
Springer Verlag | 2022
|Multiscale Object Detection from Drone Imagery Using Ensemble Transfer Learning
DOAJ | 2021
|