[Question] How to detect and grab a grid from an image

[yozachar]

Hi, I'm new to this field, so as part of my studies I'm trying to detect and return a Sudoku grid from and image. I know I can use the Hough lines transform but it gets messy:

It possible to detect grid more cleanly and efficiently?

[adeptflax]

I'm been trying figure out how to do image2image translation with this repo for the past few days. I think I figured it out. I'll write a guide on how to do this once I actually successfully trained a model. My issue #51.

[adeptflax]

@joe733 you could probably do this with pix2pixHD. Maybe CNN to lstm would be better for your use case. Though I dunno.

[adeptflax]

I did on that issue how to do it a few days ago #51.

[yozachar]

Nice, I'll go through it.

[adeptflax]

I realized I didn't do it correctly. I explained more on the issue thread. Pix2pixHD is better than the way I did it.

[adeptflax]

You would have to train a vqgan model on Sudoku images and train another vqgan model on Sudoku line patterns. You would then train a transformer that would convert different the 2 images.

[yozachar]

Hi, I'm still a beginner, so figuring things slowly. Thank you so much for the pointers!