Smart CCTV Using Deep Learning

Project Overview

Transform traditional surveillance systems with the power of Deep Learning to enhance security and safety. This Smart CCTV system incorporates real-time object detection, facial recognition, anomaly detection, automated alerts, and advanced analytics.

Key Features

1. Real-time Object Detection:

   • Utilizes state-of-the-art Deep Learning models for accurate and efficient object detection.
   • Provides real-time analysis of video feeds to identify and track objects.

2. Facial Recognition:

   • Implements facial recognition technology for identifying individuals in the surveillance footage.
   • Enhances security by allowing for the identification of known individuals.

3. Anomaly Detection:

   • Detects unusual patterns or behaviors in the monitored area.
   • Raises alerts for potential security threats or abnormal activities.

4. Automated Alerts:

   • Sends automated alerts in real-time when suspicious activities or objects are detected.
   • Improves responsiveness by notifying security personnel promptly.

5. Advanced Analytics:

   • Generates comprehensive analytics reports for historical data analysis.
   • Provides insights into patterns, trends, and potential security risks.

6. Motion Prediction Improvement:

   • Achieved an 86.7% accuracy improvement in motion prediction using the Haar Cascade algorithm.
   • Enhances the system's ability to predict and track moving objects with higher precision.

Technology Stack

• Programming Language: Python
• Web Framework: Django
• Computer Vision Library: OpenCV
• Object Detection Algorithm: Haar Cascade
• Databases: MongoDB, MySQL
• Video Streaming Protocol: RTSP
• Email Notification: SMTP
• Web Technologies: HTML, CSS, Javascript
• Cloud Platform: Azure

Installation

1. Clone the repository:

   https://github.com/SHAMSUNDAR20/Smart-CCTV-Using-Deep-Learning.git
   cd Smart-CCTV-Using-Deep-Learning