Downloads: 3 | Views: 80 | Weekly Hits: ⮙1 | Monthly Hits: ⮙1
Student Project | Computer Science & Engineering | India | Volume 12 Issue 11, November 2023
Weapon Identification using YOLO V5 Algorithm
Divyamsh Reddy A | M Poojitha [2] | G Puspalalitha | M Vishnu Vardhan Reddy | K Ashok Kumar | G Nithya Sree
Abstract: Rapid advancement of computer vision technologies has led to significant progress in the field of object detection, with applications ranging from surveillance to autonomous vehicles. One critical application is the identification of weapons in real-time scenarios, such as security monitoring and law enforcement. This paper presents a novel approach for weapon identification using the YOLOv5 (You Only Look Once version 5) algorithm, a state-of-the-art real-time object detection model. Traditional methods of weapon identification often rely on manual intervention or limited rule-based systems, which can be time-consuming and prone to errors. In contrast, the proposed method harnesses the power of deep learning to automate the detection process. YOLOv5, known for its speed and accuracy, is employed to detect and classify weapons in images and video streams. The dataset used for training encompasses a diverse range of weapon types, orientations, and lighting conditions to ensure robustness. The methodology involves fine-tuning the YOLOv5 architecture on the weapon dataset, optimizing hyperparameters, and leveraging data augmentation techniques to improve model generalization. The resulting trained model demonstrates remarkable proficiency in real-time weapon detection, outperforming traditional methods and achieving high precision and recall rates. Experimental results on benchmark datasets and custom video sequences showcase the effectiveness of the proposed approach. The YOLOv5-based weapon identification system exhibits the capability to swiftly and accurately detect weapons, thus holding significant promise for enhancing security measures in public spaces, transportation hubs, and high-security areas. In conclusion, this paper contributes to the domain of automated security systems by introducing a robust and efficient solution for weapon identification. The utilization of YOLOv5 demonstrates the potential of deep learning to address critical real-world challenges, paving the way for safer environments through advanced object detection techniques.
Keywords: Weapon, YOLO V5, Algorithm
Edition: Volume 12 Issue 11, November 2023,
Pages: 550 - 564