NextVision, an Intelligent Video Surveillance System Based on Computer Computer Vision and Natural Language Processing

Ensuring security is essential in today’s world, whether it’s to protect property or ensure personal safety. However, traditional surveillance systems initially designed to meet these needs have certain shortcomings. Personal security personnel responsible for constantly monitoring screens may miss crucial sequences. Determining when to record camera feeds to minimize storage wastage can be complex. Also, making these recorded sequences usable and actionable represents an additional challenge. Our goal is to enhance video surveillance tasks through artificial intelligence. In this paper, we introduce NextVision, a user-friendly and intelligent video surveillance platform. We combine computer vision and natural language processing to automatically identify objects in real time and provide interaction based on textual queries. Powered by a Django-based backend architecture and a frontend supported by Next JS, synchronized via sockets in real time, NextVision evolves machine learning to improve security for everyone proactively.