A Python library for extracting GPS data / images from Viofo dashcam videos.
- a GPU
- ffmpeg
- pip
sudo apt-get install ffmpegpip install -r requirements.txtInstall the CUDA toolkit:
https://developer.nvidia.com/cuda-downloads
This may require you to change your nvidia driver versions. CUDA is needed for running the OCR model on the imagery.
Clone the wcedmisten/osm2pgsql-docker repo, and follow the steps from the README.
The entire pipeline does the following:
- extract the GPS track from a dashcam clip
- extract corresponding images for each point
- run OCR on each image
- if a "speed limit" sign is found in the image, check OSM to see if the road has "maxspeed" set
- if a missing speed limit is found, output a link to the image, and a link to the OSM feature
python src/pipeline.py /path/to/input.MP4python3 -m unittest test/test.py
https://sergei.nz/extracting-gps-data-from-viofo-a119-and-other-novatek-powered-cameras/