| _version_ | 1866901258171842560 |
|---|---|
| author | Prof.Sanjay Pal, Kanish Singh, Harsh Katre, Harsh Singh Bhadoriya, Navneet Pal |
| author_facet | Prof.Sanjay Pal, Kanish Singh, Harsh Katre, Harsh Singh Bhadoriya, Navneet Pal |
| contents | <div> <div>This study introduces a real-time intruder detection system designed to make security monitoring smarter and more reliable. The system combines the strengths of classical computer vision with modern deep learning techniques. For quick and efficient face detection, it uses the Haar Cascade classifier, which helps identify individuals as soon as they enter a monitored area. To go beyond simple recognition, the framework integrates an LRCN (Long-term Recurrent Convolutional Network) model that analyzes behavior over time, spotting unusual or suspicious activities by learning spatiotemporal patterns. By blending frame-based facial recognition with sequencebased activity analysis, the system achieves robust intruder detection while keeping false alarms to a minimum. It works in real time, automatically sending alerts whenever abnormal events occur, and is flexible enough to be deployed in both indoor and outdoor environments. Experimental results show that this hybrid approach— mixing traditional methods with deep learning— significantly boosts accuracy, responsiveness, and overall reliability. As a result, the system proves to be a strong candidate for modern intelligent surveillance infrastructures.</div> </div> |
| format | Recurso digital |
| id | zenodo_https___doi_org_10_5281_zenodo_18160785 |
| institution | Zenodo |
| language | |
| publishDate | 2026 |
| publisher | Zenodo |
| record_format | zenodo |
| spellingShingle | Advanced Real-Time Intruder Detection System Using Multimodal Deep Learning: Integrating Visual, and Behavioral Analysis Prof.Sanjay Pal, Kanish Singh, Harsh Katre, Harsh Singh Bhadoriya, Navneet Pal <div> <div>This study introduces a real-time intruder detection system designed to make security monitoring smarter and more reliable. The system combines the strengths of classical computer vision with modern deep learning techniques. For quick and efficient face detection, it uses the Haar Cascade classifier, which helps identify individuals as soon as they enter a monitored area. To go beyond simple recognition, the framework integrates an LRCN (Long-term Recurrent Convolutional Network) model that analyzes behavior over time, spotting unusual or suspicious activities by learning spatiotemporal patterns. By blending frame-based facial recognition with sequencebased activity analysis, the system achieves robust intruder detection while keeping false alarms to a minimum. It works in real time, automatically sending alerts whenever abnormal events occur, and is flexible enough to be deployed in both indoor and outdoor environments. Experimental results show that this hybrid approach— mixing traditional methods with deep learning— significantly boosts accuracy, responsiveness, and overall reliability. As a result, the system proves to be a strong candidate for modern intelligent surveillance infrastructures.</div> </div> |
| title | Advanced Real-Time Intruder Detection System Using Multimodal Deep Learning: Integrating Visual, and Behavioral Analysis |
| url | https://doi.org/10.5281/zenodo.18160785 |