Video page scanner

Simple video page scanner that applies an homography in all video frames of the video and joins them.
It receives a video containing a paper (example) and outputs a video only with the paper changing along with the original video. This project was made for the Computer Vision course.

Running

Install dependecies

pip3 install -r requirements.txt

Run

python3 main.py -v VIDEO -o OUTPUT -t TEMPLATE

• VIDEO - Video input file containing the paper (default: video.mp4)
• OUTPUT - Name of the output video file (default: output.mp4)
• TEMPLATE (optional) - Image file containing the wanted dimensions of output paper in the video (default dimensions: 1275x1650)

Example

You can run the program with the example video file

video.mp4

• Expected output

output.mp4

Image preview example

Input	Output
)

Others

• Original project statement: statement.pdf