International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 10 Issue: 07 | July 2023
p-ISSN: 2395-0072
www.irjet.net
Social Distance Monitoring and Mask Detection Using Deep Learning Techniques M Sneha1, Prasad A M2 1Student, M.Tech, Computer Science Engineering, Dayananda Sagar College of Engineering
2Professor, Department of Computer Science Engineering, Dayananda Sagar College of Engineering
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - The proposed system uses a combination of
distancing compliance in various public environments. The development of a real-time social distancing system using image recognition technology and his YOLOv5 algorithm is expected to help curb the spread of infectious diseases like COVID-19. The proposed system provides a cost-effective, accurate and efficient solution for enforcing social distancing guidelines in various public settings by automating the surveillance process. Its applications span multiple areas such as public health, workplaces, events, transportation, retail and education, making it a valuable tool for promoting safety and mitigating the impact of pandemics and infectious diseases. As technology continues to advance, such systems could play an important role in shaping a safer and more secure future for society.
image processing and machine learning techniques to analyze video feeds and image sequences from surveillance cameras and other visual sources. Accurately and efficiently identify people in a scene using the YOLOv5 algorithm, a state-of-the-art deep learning object detection model. By analyzing the spatial relationships between detected individuals, the system can infer whether social distancing guidelines are being followed. This methodology includes several critical steps. The system then uses image preprocessing techniques to improve the quality of the input image and extract meaningful features. Combining background subtraction with foreground segmentation can help identify areas of interest populated by people. These regions are input into his YOLOv5 model for object detection. The system uses a geometric analysis approach to determine if social distancing is being followed. By estimating the distance between each pair of detected individuals, the interpersonal distance is calculated and compared to a pre-defined social distance threshold. Violations are reported and displayed in real time so that relevant authorities are notified in a timely manner.
1.1 PROBLEM STATEMENT This issue explores the need for automated systems to enforce social distancing protocols in public spaces during the COVID-19 pandemic. The proposed system aims to detect adherence to social distancing guidelines in real time using image recognition technology and his YOLOv5 algorithm. By analyzing video feeds and image sequences from surveillance cameras, the system can identify people and assess their spatial relationships to determine if social distancing is being followed. The proposed system aims to detect adherence to social distancing guidelines in real time using image recognition technology and his YOLOv5 algorithm. By analyzing video feeds and image sequences from surveillance cameras, the system can identify people and assess their spatial relationships to determine if social distancing is being followed.
Key Words: YoloV5 algorithm, social distancing, image processing, real-time detection
1. INTRODUCTION The outbreak of the novel coronavirus disease (COVID-19) in 2019 has had a major impact on societies around the world. To curb the spread of the virus, health officials and governments have taken various measures, including social distancing. The Objective of social distancing is to reduce close physical interactions between individuals, thereby curbing transmission of the virus. Social distancing in public places is essential to ensure public safety and stop the spread of the pandemic. However, manually monitoring compliance with social distancing guidelines can be labor-intensive, time-consuming, and error-prone. To meet this challenge, researchers and engineers have developed automated social distancing detection systems using advanced technologies such as image recognition and deep learning algorithms. This research focuses on utilizing image recognition technology and his YOLOv5 algorithm, a state-of-the-art deep learning model, to enable accurate real-time monitoring of social
© 2023, IRJET
|
Impact Factor value: 8.226
1.2 OBJECTIVES 1) Development of real-time social distance detection system: The main purpose of this research is to design and develop a real-time social distance detection system using image recognition technology and his YOLOv5 algorithm. The system processes video feeds or series of images from surveillance cameras to identify people at crime scenes. 2) Fine Tuning of YOLOv5 Social Distance Detection Algorithm: To achieve accurate and efficient social distancing detection, the YOLOv5 algorithm has been improved using a dedicated dataset created specifically for this purpose. This dataset contains a variety of
|
ISO 9001:2008 Certified Journal
|
Page 1057