zxing-js/library

[CODE_128] Can't recognize long values with mixed 0-9A-z content

ashvejnov opened this issue · 3 comments

Hi there,

I'm a bit in trouble, I can recognize numeric CODE_128 barcodes with any lenght, but when I face long CODE_128 BC with mixed content - it starts to fail when total lenght of content exceeds 15 characters.

Examples:

BJ0WX0E9500V1NAQ0186 - my goal, can't be recognized at all (20 characters).

BJ0WX0E9500V1NA - can be recognized pretty well (15 characters and shorter)

BJ0WX0E9500V1NAQ - still can be recognized, but I see chaotic random special characters in result, something goes wrong starting from here (16 characters)

BJ0WX0E9500V1NAQ0 - can't be recognized completely (17 characters and longer)

Pls kindly dive deep into this bug.

Appreciate your great game changing work!!

UPDATE: Fixed this issue by enlarging video box size in capturing UI... yeah) but still, it seems to be weird a bit, that decoding is so depending on UI viedeo frame/box size.

I would be very interested on knowing how to improve barcode 128 reading capabilities.

If you could elaborate what you mean with video box size I would be very grateful.

Doesn't matter how hard I try, even with perfect contrast imagines like the one attached doesn't seem to work.

Example:

https://ibb.co/hWkGHzm

Stale issue message