The Human Detection in Surveillance Cameras is a real-time project that detects humans in video streams using the Histogram of Oriented Gradients (HOG) algorithm. When a human is detected, the system captures an image and sends an email alert. This project is useful for enhancing security in public spaces, monitoring pedestrian traffic, and preventing accidents.
- Real-Time Detection: The system continuously analyzes video frames from a camera feed and identifies pedestrians in real time.
- Email Alert: Upon detecting a human, the system captures an image and sends an email notification to a specified recipient.
- Customizable Parameters: You can adjust detection parameters (e.g., window stride, padding, scale) to fine-tune the system's performance.
- Minimal False Positives: The HOG-based approach minimizes false positives, making it suitable for practical applications.
- The HOG algorithm computes gradient histograms in local image regions.
- It represents the shape and texture of objects by analyzing the distribution of gradient orientations.
- HOG is particularly effective for detecting pedestrians due to its robustness against variations in lighting and pose.
-
Clone the Repository:
git clone https://github.com/Dileep62892/Human-detection-in-surveillance-cameras.git
-
Install Dependencies:
- Python 3.x
- OpenCV
- smtplib (for email notifications)
pip install opencv-python imutils
-
Configure Email Settings:
- Open human.py.
- Set your Gmail address and password.
- Update the recipient’s email address (toadd).
-
Run the human detection script:
python human.py
-
Press ‘q’ to exit the program.