Skip to content

Robust color handling via zbar 8bpp conversion #170

Open
NaziaAfreen015 wants to merge 2 commits into
NaturalHistoryMuseum:masterfrom
NaziaAfreen015:master
Open

Robust color handling via zbar 8bpp conversion #170
NaziaAfreen015 wants to merge 2 commits into
NaturalHistoryMuseum:masterfrom
NaziaAfreen015:master

Conversation

@NaziaAfreen015

Copy link
Copy Markdown

While pyzbar.decode converts PIL image into greyscale or enforces greyscale image for a (pixel, width, height) argument; it currently grabs only the first channel from NumPy arrays. That works for grayscale/standard RGB, but fails when the QR foreground/background is carried in non-first channels (e.g., red-only scenes). Since libzbar already provides reliable color→8bpp conversion, this PR routes array inputs through zbar’s converters to ensure correct decoding regardless of which channel holds the signal.

Add thin wrappers for zbar_image_convert, zbar_image_convert_resize, zbar_image_get_data, and zbar_image_get_data_length.

When a NumPy array with 3/4 channels is passed, convert to 8bpp via libzbar before scanning.

Preserve existing grayscale path; no behavior change for 8bpp inputs.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant