git # Video Frame Stitching for CCTV Camera Movement

This project extracts frames from a CCTV video only when the camera is moving, stitches them into a panorama when the camera stops, and saves each stitched image.

Features

• Detects camera movement using frame-to-frame difference (motion detection)
• Extracts and collects frames only during camera movement
• Stitches frames into a panorama when the camera stops
• Saves each panorama to the StitchedFrame directory
• Handles multiple movement events in a single video (creates multiple panoramas)

Requirements

• Python 3.x
• OpenCV (opencv-python)
• numpy

Install dependencies:

pip install -r requirements.txt

Usage

Run the frame processor script:

python frame_processor.py <video_path>

Optional Arguments

• --max-frames: Maximum number of frames to process per movement (default: 100)
• --output-dir: Output directory for stitched images (default: StitchedFrame)

Example

python frame_processor.py /path/to/video.mp4 --max-frames 100 --output-dir StitchedFrame

The stitched images will be saved in the StitchedFrame directory as stitched_panorama_1.jpg, stitched_panorama_2.jpg, etc.

Output

• Each panorama image is saved in the StitchedFrame folder, one for each camera movement event.

Notes

• The script uses motion detection to determine when the camera is moving.
• The quality of stitching depends on the video and camera movement.
• You can adjust the motion detection threshold in the MotionDetector class in frame_processor.py if needed.

Troubleshooting

• If stitching fails, check the log for errors. Not enough overlap or too few frames can cause failure.
• For best results, use videos with smooth camera movement and good lighting.