INTELLIGENT VIDEO SURVEILLANCE
USING DEEP LEARNING
• Final Year Project
• Using Deep Learning Techniques
Abstract
• Detects abnormal activities in surveillance
videos
• Uses Deep Learning (Spatial Autoencoder)
• Supports real-time and uploaded video
detection
• Abnormal frames saved for review
Introduction
• Domain: Artificial Intelligence & Computer
Vision
• Used for security monitoring
• Automates abnormal activity detection
• Overcomes limitations of manual surveillance
Problem Statement
• Manual surveillance is inefficient
• Delayed response to abnormal events
• High human effort required
• Need for automated intelligent system
Existing System
• Manual video monitoring
• Requires continuous human attention
• Prone to errors
• Time-consuming
Disadvantages of Existing System
• Low efficiency
• High manpower required
• Delayed detection
• Not scalable
Proposed System
• Automated video surveillance
• Deep learning-based detection
• Detects violence or abnormal behavior
• Stores abnormal frames automatically
Advantages of Proposed System
• Real-time detection
• High accuracy
• Reduced human effort
• User-friendly interface
Modules
• User Module
• Video Upload Detection
• Real-Time Camera Detection
Technologies Used
• Frontend: HTML, CSS, JavaScript
• Backend: Python (Django)
• Database: MySQL
• Algorithm: Spatial Autoencoder
Hardware Requirements
• Intel i3 or higher
• 4GB RAM or more
• 100GB Hard Disk
Software Requirements
• Python
• Django Framework
• XAMPP Server
• Sublime Text Editor
Applications
• Public surveillance systems
• Banks and ATMs
• Shopping malls
• Security monitoring
Conclusion
• Efficient abnormal activity detection
• Reduces human effort
• Improves security
• Suitable for real-time deployment
Future Scope
• Cloud deployment
• Mobile application
• AI model enhancement
• Multi-camera support
Thank You
• Thank You
• We are happy to answer your questions

Intelligent_Video_Surveillance_Project_PPT.pptx