International Journal of Innovative Research in Computer and Communication Engineering
ISSN Approved Journal | Impact factor: 8.771 | ESTD: 2013 | Follows UGC CARE Journal Norms and Guidelines
| Monthly, Peer-Reviewed, Refereed, Scholarly, Multidisciplinary and Open Access Journal | High Impact Factor 8.771 (Calculated by Google Scholar and Semantic Scholar | AI-Powered Research Tool | Indexing in all Major Database & Metadata, Citation Generator | Digital Object Identifier (DOI) |
| TITLE | A Review of Edge AI for Autonomous Target Detection in UAVs using NVIDIA Jetson Orin |
|---|---|
| ABSTRACT | This paper presents a practical approach to integrating Edge AI in unmanned aerial vehicles for autonomous target detection for defense applications using the NVIDIA Jetson Orin platform. The pro- posed system leverages a convolutional neural network-based deep learning model, an optimized version of YOLOv8, deployed directly on the Jetson Orin for real-time onboard inference. UAV-mounted cam- eras capture continuous aerial video feeds that are processed locally using hardware-accelerated infer- ence via TensorRT, removing dependence on cloud infrastructure and reduction of latency. The model trained on both publicly available and custom preprocessed and curated military-relevant datasets en- suring accurate target classification under various environmental conditions. Experimental results show that the system achieves more than 90% detection accuracy with an inference speed over 30 fps per sec- ond, maintaining consistent performance during flight. The research contribution towards a lightweight, energy-efficient, and deployable solution for autonomous surveillance, enhancing the decision making competency and operational efficiency of UAVs in tactical defense applications. |
| AUTHOR | HIMANSHU MAHENDRAKUMAR TRIGUNE, ASWAD DIPAK GORIVALE, PROF. PRAWIN GAWANDE |
| VOLUME | 176 |
| DOI | DOI: 10.15680/IJIRCCE.2025.1311054 |
| pdf/54_A Review of Edge AI for Autonomous Target Detection in UAVs using NVIDIA Jetson Orin.pdf | |
| KEYWORDS |
