Integrated Approach: AI-based irregular action detection using advanced drone surveillance within C2ISR systems.

Muhammad ALI

Diplomová práce

Abstract:

The rising complexity of public safety concerns needs creative techniques for observation and response, especially at large public gatherings. This thesis present a sophisticated system architecture for a Command, Control, Intelligence, Surveillance, and Reconnaissance (C2ISR) system, integrating advanced artificial intelligence (AI) with advanced drone technology to enhance the detection of irregular actions by people. The primary purpose is to design an integrated system architecture that provides real-time, AI-enabled detection of irregular actions through drone-captured video streaming, employs advanced AI algorithms for rapid analysis, and provides precise geolocation information. The development of the C2ISR system architecture utilizes unmanned aerial vehicles (UAVs), edge and cloud computing, and sophisticated AI algorithms. The lack of a sufficient irregular action dataset on open-source AI platforms prompted the construction of a custom dataset, which focuses on specific irregular activities like running, falling, fighting, and unauthorized admission into restricted zones. This thesis discusses the selection and development of state-of-the-art real-time AI detection models that were trained on a collected dataset acquired utilizing drone imagery in a controlled environment. This thesis carefully examined and validated the models' efficacy through both laboratory and field studies. The findings of this investigation revealed the system's ability to reliably and efficiently recognize and evaluate irregular actions, considerably boosting situational awareness and facilitating rapid response coordination. At the end of the thesis, there is a discussion about how to create scalable, adaptive, and morally sound AI-driven C2ISR systems in the future, with the goal of setting new standards in public safety technology. This work emphasizes the potential for combined AI and drone technology to improve public safety operations, enabling more rapid and effective monitoring and intervention techniques.

Abstract:

The rising complexity of public safety concerns needs creative techniques for observation and response, especially at large public gatherings. This thesis present a sophisticated system architecture for a Command, Control, Intelligence, Surveillance, and Reconnaissance (C2ISR) system, integrating advanced artificial intelligence (AI) with advanced drone technology to enhance the detection of irregular actions by people. The primary purpose is to design an integrated system architecture that provides real-time, AI-enabled detection of irregular actions through drone-captured video streaming, employs advanced AI algorithms for rapid analysis, and provides precise geolocation information. The development of the C2ISR system architecture utilizes unmanned aerial vehicles (UAVs), edge and cloud computing, and sophisticated AI algorithms. The lack of a sufficient irregular action dataset on open-source AI platforms prompted the construction of a custom dataset, which focuses on specific irregular activities like running, falling, fighting, and unauthorized admission into restricted zones. This thesis discusses the selection and development of state-of-the-art real-time AI detection models that were trained on a collected dataset acquired utilizing drone imagery in a controlled environment. This thesis carefully examined and validated the models' efficacy through both laboratory and field studies. The findings of this investigation revealed the system's ability to reliably and efficiently recognize and evaluate irregular actions, considerably boosting situational awareness and facilitating rapid response coordination. At the end of the thesis, there is a discussion about how to create scalable, adaptive, and morally sound AI-driven C2ISR systems in the future, with the goal of setting new standards in public safety technology. This work emphasizes the potential for combined AI and drone technology to improve public safety operations, enabling more rapid and effective monitoring and intervention techniques.

Keywords

Public Safety, Command, Control, Intelligence, Surveillance, and Reconnaissance (C2ISR), Artificial Intelligence (AI), Unmanned Aerial Vehicles (UAVs), Drones, Real-time Detection, Irregular Actions, Video Streaming, Edge Computing, Cloud Computing, AI Algorithms, Public Safety Technology

Jazyk práce: angličtina

Datum vytvoření / odevzdání či podání práce: 20. 8. 2024

Obhajoba závěrečné práce

• Vedoucí: prof. Dr. Helena Liebelt

JIHOČESKÁ UNIVERZITA V ČESKÝCH BUDĚJOVICÍCH

Přírodovědecká fakulta

Magisterský studijní program / obor:
Artificial Intelligence and Data Science / Artificial Intelligence and Data Science