Analysing the Effectiveness of YOLO Model in Detecting the Images Captured by the Drone

The traditional forensic methods are not adequate to meet the demands of forensics investigations of drone. This is mainly due to unique characteristics of drones, there are gaps in current methodologies, standards, and demonstration of fundamental issues in terms of forensics analysis. This research discusses the applications of Machine Learning models: Convolutional Neural Network (CNN) and You Only Look Once (YOLO) in the field of drone forensics and the efficiency of crime investigations in automatically detecting and classifying drones. The fast emergence of drone technologies demonstrates the urgent need to pay attention to subsequent impacts in forensic science, especially in relation to the huge numbers and wide variety of drones that make it impossible to process using traditional forensic methods. The main objective of this work is directed toward the development of a computer vision-based framework for the enhancement of the accuracy and the speed of forensic analysis for processing complex, as well as large data derived from drones. At the core of this attempt is, therefore, the development of a detailed architecture for drone forensics, further followed by a painstaking evaluation of the efficacy of CNN and YOLO models in enhancing the analysis of crime scenes and interpretation of pieces of evidence.