MedicineBank — PDF Image Studio

Drop source lecture PDF here or click to browse No file chosen

Extraction Mode Skip icons smaller than (px) Page render quality

AI Vision Labeling & Smart Filtering (Step 1)

Auto-label images with a short medical description (used for ZIP filenames and Step 2 textbook matching) Flag non-medical / useless images (pure text slides, logos, decorative shapes, blank slides) & auto-detect recurring template graphics Automatically mark AI-flagged images as removed (you can restore any of them below)

Drop target textbook PDF (with green placeholders) here or click to browse No file chosen

Placeholder Color Tolerance Minimum Box Size (% of page width) How Replaced Images Fit Export PDF Quality & Size

Erase empty green boxes with background color on export Force pure white (#FFFFFF) mask instead of adaptive page background color

AI Auto-Fill Entire File (Global Chapter Matching + Deep Visual Verification)

Full-File Automation Mode Minimum AI confidence to place

Only fill empty placeholders (keep images you already placed) Allow the same extracted figure to be used in more than one placeholder Include text from the image's source lecture slide when matching Auto-label Step 1 figures that have neither an AI label nor slide text before matching (saves 90% of API quota)

1-Click Full-File Automation: Evaluates all placeholders across the entire PDF automatically (no need to click “AI Pick” on each placeholder individually), verifies ambiguous figures visually, and provides 1-click runner-up alternatives on every slot.

PDF Image Studio v2.0 · AI Vision & Textbook Layout Engine

Live Page Preview (click any green box to assign; toggle “Draw / Edit Boxes” to add/resize/delete boxes)