[Question] Image processing resources for paper data storage

[Linguofreak]

I'm mulling over the idea of coding up a reader for QR-style pixel-grid barcodes. This would be a retrocomputing project with the idea of printing out barcodes where the pixel arrangement of the central part of the barcode would resemble legacy punch card / punch tape hole layouts, with metadata, checksum, and alignment/registration features (plus maybe some human-readable text) in the periphery, and the idea is that these could then be read back with a smartphone app and sent to a listening retrocomputing emulator listening for a telnet connection on some port.

Since a certain aesthetic is desired, I'd rather not use plain QR codes, for which appropriate libraries are a dime a dozen.

Does anybody know of any of the following resources:

1) Existing open-source barcode libraries that can read in a user specification for a 2D barcode format? Essentially, the user defines a pixel grid of a certain size with certain alignment features, the library finds the alignment features in a camera image, determines the orientation and size of the barcode in the image, and then uses that to fill in bit values in the pixel grid?
2) Existing libraries with hard-coded formats where different parts of the process are abstracted out from each other well enough in the code that a novice to the field could hope to code up a new format?
3)Tutorial resources that cover the knowledge necessary to code up 1) or 2) from scratch?

 

[kuddel]

I see your idea, but it might not be that easy.

In a quite similar project we just "faked it" by sticking thin RF-ID Tag stickers onto the punchcards.
The reader then just needed to read the ID and from a lookup "database" spit out the stored data[1]; and there we had it.
With an arduino, an RFID reader, an SD-Card for the "lookup database" and some minor programming skills this was working 'better than expected'.

[1] ... a complete binary in that case - which would never have fitted onto the punchcard anyhow