I've brought Speech Bubble Density Estimator to Gold with v2.0.0.
This one looks through local comic page images or a CBZ and scores areas with light backgrounds and dark, high-contrast marks. That can help point you toward pages or stretches that may be heavier on dialogue, or ones that give more space to the art.
You can adjust the thresholds, review ranked pages, and generate a contact sheet showing the candidate areas. It doesn't read the words. White artwork, captions, and unusual lettering can affect the results, so being able to see what contributed to the score matters.
The current release requires Python 3.11+ and uses Pillow 13. This Gold pass also added tighter per-page limits and clearer reporting when some pages couldn't be analyzed. Everything stays local, with no OCR service or image upload involved.
https://t.co/7caRCtL6UK
I've brought Speech Bubble Density Estimator to Gold with v2.0.0.
· Other posts