mkingyixing/AI-QR-code-scanning-

QR and barcode decoding for DSH from image files, the live screen or a webcam, plus QR generation — all offline.

Gives the model eyes for codes. qr_scan decodes every code in an image — a path, a glob or an http(s) URL — and returns positions with each payload, qr_scan_screen captures the screen or a region and decodes it, qr_scan_camera grabs a single webcam frame, and qr_make renders text or a URL as a QR PNG in the workspace. Each result is classified by kind (url, wifi, contact, email, phone, sms, numeric, text), so the model can act on a scanned Wi-Fi credential or contact card without guessing from the raw string. Output directory, maximum input size and timeout are configurable per row, and screenshots stay on the machine.

Vision & Multimodal ★ 0
View on GitHub ↗

Install

dsh plugin --profile web add github:mkingyixing/AI-QR-code-scanning-

The README's recommended install, straight from GitHub (the repository name ends with a hyphen, and that is the literal spec). It is not published to npm: the README explains npm rejects uppercase package names, so publishing would need a lowercase name. A local PowerShell install script is documented for the desktop profile, with -Uninstall to remove.

Compatibility

Decoding runs on zxing-cpp compiled to WebAssembly (zxing-wasm), in-process, so there is no interpreter startup. The README's measurements on its own 10-image corpus — rotation, inversion, a 78px thumbnail, blurry JPEG, perspective and two codes in one frame — are 10/10 decoded, 92 ms cold and 13.6 ms warm per decode against ~750 ms for the equivalent Python script, and 227 ms for a full-screen capture plus decode against ~1 500 ms. Screen capture uses GDI and camera frames go through ffmpeg; everything is local unless you pass an explicit http(s) URL to qr_scan.

Details

Alternatives

kaixinbaba/dsh-vision-recognizer · siegfly/dsh-deepseek-vision · shinegeer/dsh-vision-window

More plugins in Vision & Multimodal

Browse more in Vision & Multimodal

Guides for Vision & Multimodal plugins