AI-Powered Tool

Turn any fish photo into a sticker

AI background removal, automatic species recognition, and clean white-outlined stickers — all processed locally on your machine.

✂️

AI Background Removal

🔍

Species Recognition

macOS Vision OCR reads labels from composite images and matches against 144 species in our database.

🎨

Sticker Rendering

White outline + gray accent ring on transparent background. Works beautifully on any theme.

📐

Composite Splitting

Upload a grid of labeled fish — FishPic detects each cell, crops, and processes them individually.

↔️

Orientation Control

Flip fish to face left (standard for species illustrations) with a single click before saving.

🗂️

Gallery Management

Browse, search, and manage all processed species with cutout and original photo support.

How the pipeline works

Upload

Drag & drop a single fish photo or a composite grid image into the browser interface.

Detect & Crop

For composite images, OCR detects grid labels and crops each fish cell. Single photos pass through directly.

Background Removal

rembg (U2-Net) removes the background, producing a clean alpha-channel cutout of the fish body.

Sticker Rendering

White ring + gray accent ring are computed via alpha expansion and gaussian blur, composited on transparent background with cream interior fill.

Species Matching

OCR text or manual input is resolved against 144 species — FishPic suggests the best match and auto-fills metadata.

Save to Gallery

One click saves the 1200×1200 lossless WebP into the species photo library, instantly visible in the gallery.

FishPic runs as a local web server on your Mac. Requires Python 3.10+ and macOS (for Vision OCR).

python3 fishpic.py

Click to copy · Opens at localhost:19527