Litcius/Paper detail

DCNN Based Activity Classification of Ornithopter using Radar micro-Doppler Images

Aditya Srinivas Akella, A. Arockia Bazil Raj

202116 citationsDOI

Abstract

This research work has proposed a method to classify different flight modes of an ornithopter. A custom- made Continuous Wave (CW) Radar operated in X-Band at 10 GHz has been used to achieve this. An ornithopter is operated in an indoor environment in various flight modes, and the Doppler signatures are collected. These Doppler signatures are then converted as images. A series Deep Convolutional Neural Network (DCNN) was built and trained rigorously and tested on these collected Doppler images. Our proposed network has achieved an accuracy of 97% after testing.

Topics & Concepts

RadarDoppler radarComputer scienceConvolutional neural networkArtificial intelligenceRemote sensingDoppler effectRadar imagingComputer visionGeologyPhysicsTelecommunicationsAstronomyFish biology, ecology, and behaviorAnimal Behavior and ReproductionWater Quality Monitoring Technologies
DCNN Based Activity Classification of Ornithopter using Radar micro-Doppler Images | Litcius