Enforcement of CNN Model in Drone Detection System

Drones play a crucial role across various applications. However, their widespread use has raised concerns about potential misuse, includi surveillance, solicitation, and terrorism. Consequently, there is a growing demand for effective drone detection systems. A promising solution for this is the application of Convolutional Neural Networks (CNNs). Rhynchus is a system designed specifically for image recognition tasks. CNNs have proven highly effective in detecting drones in images, even under challenging conditions such as thick fog. In this paper, the objective is to develop a drone detection system with various applications. The proposed system consists of two main components: a feature extraction module and a classification module. The feature extraction module identifies relevant characteristics, such as shape, texture, and movement, from input images. The classification module uses these features to classify an object as either a drone or a non-drone. We will utilize a large dataset of drone images for training and evaluating the system, which includes a diverse array of drones and environments, ensuring robustness across different scenarios. Once developed, the system will be implemented on a real-time platform to showcase its effectiveness in real-world settings. It will be capable of detecting drones in real time using video feeds from cameras. The proposed drone detection system offers several advantages: it is accurate and resilient under various conditions, cost-effective to implement, and scalable for use in a wide range of scenarios.