SuperRes-OpenCV

Super Resolution by OpenCV

References

• https://learnopencv.com/super-resolution-in-opencv/
• https://towardsdatascience.com/deep-learning-based-super-resolution-with-opencv-4fd736678066
• https://www.youtube.com/watch?v=-LqHr5V67C4&t=586s

Procedures

1. create the virual env in a blank folder
   • python -m venv SuperRes
   • cd SuperRes
   • Scripts\activate
2. install dependencies (https://pyimagesearch.com/2018/09/19/pip-install-opencv/)
   • pip install opencv-contrib-python (recommended)
3. download your needed models
   • https://learnopencv.com/super-resolution-in-opencv/ (read model section, the links are there)
4. modify as you needed and run the codes
   • modify the upscale.py and main() function as you needed
   • run by the command, python upscale.py

Google Colab Demo

here

Models

There are currently 4 different SR models supported in the module. They can all upscale images by a scale of 2, 3 and 4. LapSRN can even upscale by a factor of 8. They differ in accuracy, size and speed.

1. EDSR [1].
   • This is the best performing model.
   • However, it is also the biggest model and therefor has the biggest file size and slowest inference.
   • You can download it here.
2. ESPCN [2].
   • This is a small model with fast and good inference.
   • It can do real-time video upscaling (depending on image size).
   • You can download it here.
3. FSRCNN [3].
   • This is also small model with fast and accurate inference.
   • Can also do real-time video upscaling.
   • You can download it here.
4. LapSRN [4].
   • This is a medium sized model that can upscale by a factor as high as 8.
   • You can download it here.