JubJub — Product Image Grabber

🛍 All-in-one image, text and color extraction tool for people who work on the web JubJub automatically finds product images and videos on any web page, recommends and organizes them, and supports copy / save / background removal / AI upscale / annotation / video and GIF frame editing / area capture / full-page capture / multi-area capture / area GIF recording / text extraction (OCR) / color extraction / find similar images / basket, all in one browser extension. --------------------------------------------------------------------------------------------------------- ✨ Main features ▼ Automatic image scan ( default: Alt + Shift + J [ change it in Manage Extensions > Keyboard shortcuts ] ) - Extract and organize every product image on the page in one click (including images inside iframes and CSS background images) - Automatic duplicate removal based on pHash (robust against CDN variants) - Multiple resolutions of the same image are grouped into the largest one with a +N badge (click the badge to pick or expand the other resolutions) - Per-domain learning improves accuracy on your next scan - Automatic sorting into Recommended (product solo shots) / Tall images / Others / PNG / WEBP / SVG ▼ Video grab and frame editing - Finds videos on product pages too, so you can check them by poster thumbnail and save them - Frame editing: pause at the scene you want and capture it as an image, then apply crop, background removal, annotation and upscale to it - Floating frame bar at the bottom of the screen (play/pause, previous/next, seek bar, collapsible) - GIFs can be edited frame by frame as well (pause and play, restore the original) ▼ AI upscale ⭐ NEW - Enlarge small images by 1.6× / 2× / 4× in the crop (edit) window - Fast mode (Real-ESRGAN): handles large images too, and uses every CPU core for speed - Quality mode (waifu2x): removes JPEG artifacts while keeping detail (for small images only, with progress and time remaining, cancel anytime) - Models are bundled, so it runs fully offline with no download and nothing leaves your device - The result is added as a new image, so you can compare it side by side with the original ▼ Find similar images - Paste an image from the clipboard or pick a local file, and JubJub finds similar images among the scan results - Even a partially cropped image (a screenshot from a design doc, for example) can find the original by containment - Results appear live while searching, and can be selected, saved and edited just like the main gallery ▼ Google image search - Hover a scanned thumbnail and a search button appears to the left of the crop button - One click runs a reverse image search on Google (Google Lens) - Handy for finding other sellers of the same product, a larger original, or the source ▼ Three background removal modes - ✂ Solid color removal (LAB distance, with a slider for fine tuning) - 🧠 U²-Net AI cutout (fully local, works offline). Adjust sharpness directly with the edge threshold and softness sliders in the result window - 🎯 remove.bg API precision cutout (optional, enabled when you enter your own API key) - In the subject crop gallery you can apply it to selected or all items and reset to the original. The sliders (threshold, softness, removal strength, white protection) re-render instantly without running inference again ▼ Smart auto crop - In crop (edit) mode, click ✨ Auto crop and JubJub detects subjects and proposes crop areas - Even a long product page or several products in one image are split per subject (tune it with the split strength slider) - 📐 Uniform sizing: apply a ratio (original / 1:1 / 4:5 / 16:9 / 9:16), margin and background (transparent / white / color) to every crop at once, with undo ▼ Three capture modes ( Alt + J ) - Area capture: drag to select part of the page (8 resize handles, arrow keys for 1px and Shift + arrows for 10px precision) - Full site capture: scrolls the whole page automatically and stitches it into a single PNG (sticky headers handled automatically) - Multi-area capture: scroll freely, select several areas, and save them all as one ZIP ▼ Area GIF recording and editing - 🎬 In capture mode, pick an area and press the shortcut (Alt+J) to record a GIF of on-screen motion (up to 30 seconds) - ✂ Trim while previewing in the editor: start and end trim, fps, output width, color count and dithering, loop count - 🖱 Drag on the preview to crop just that area for saving (resize and move with handles, click outside to cancel) - ⚡ Speed (0.5× to 2×) and 🔁 boomerang (forward plus reverse loop) effects ⭐ NEW - ⏯ Pause the preview with the ▶/⏸ button or the Space key, plus shortcuts for frame-by-frame movement - 💾 Save as GIF or WebM (exactly the range and area you selected). You can cancel during encoding ▼ Annotation ( in crop / edit mode ) - Draw freely on the image with pen, highlighter, line, rectangle, ellipse and arrow in the crop (edit) window - Adjust color (palette, eyedropper, color code) and thickness, and move, resize or delete what you drew - Undo and redo (Ctrl+Z, Ctrl+Y), and hold Shift for perfect squares, circles, 45° lines and straight pen strokes - Annotations are baked into copies and PNG downloads only (background removal, OCR and auto crop always use the clean original) ▼ JubJub basket ⭐ NEW - 🧺 Collect images you like across several sites in the basket, then download them all as one ZIP - In the thumbnail card grid you can click to copy, send to crop, or edit a selection (Shift click for a range) - On the main screen you can also check several images and open them all in the crop window with Edit - Kept even after you close and reopen the browser (not subject to automatic cleanup) ▼ Color picker ( Shift + I ) - Pick the color of any pixel on screen (native browser EyeDropper API) - Switch between HEX / RGB / HSL and keep an automatic color history (180 days) - Works anywhere, including outside the page (browser UI, desktop, and so on) ▼ Text extraction (OCR) - Local inference with th…