People Counting System Using Video Camera

A real-time people counting system that uses Intel RealSense camera and YOLO object detection to detect and count people in a classroom. The system also leverages accuracy and logs occupancy data for analysis.

Features

• Real-Time People Detection and Tracking:
  • Uses YOLO object detection to identify and track people in video frames.
• Depth-Based Filtering:
  • Uses depth data from RealSense cameras to filter out false positives.
• Automatic Image Capture:
  • Captures and saves combined images (RGB, IR, depth, and detection overlays) every 15 minutes.
• Data Logging:
  • Logs occupancy data (timestamp, current count, max count) to CSV files.
• Visualization:
  • Displays real-time visual feedback, including bounding boxes, counts, and detection overlays.
• Occupancy Analysis:
  • Analyzes logged data to display recent occupancy statistics and generate data tables.

Requirements

• Hardware:
  • Intel RealSense camera (or a regular webcam).
• Software:
  • Python 3.8+
  • OpenCV
  • Pyrealsense2
  • Ultralytics YOLO
  • NumPy
  • MediaPipe (optional for pose detection)

Installation

1. Clone the repository:

git clone https://github.com/yourusername/Counting-People-Using-Video-Camera.git
cd Counting-People-Using-Video-Camera

2. Install required packages:

pip install -r requirements.txt

3. Download YOLO model weights:
   • Place the YOLO model file (e.g., yolov8m.pt) in the appropriate directory.

Usage

1. Run the program:

cd src
python people_counter.py

2. Controls:

• Press 'q' to quit the program

3. Analyze logged data:
   • Use the occupancy_analysis.py script to analyze logged data and generate statistics:

     python occupancy_analysis.py