Goddy36-A/Autopass

Auto passport photo generator for Uganda - Kotlin/Compose # Autopass Uganda Native Android app (Kotlin + Jetpack Compose) that takes a photo with the in-app camera and automatically turns it into a Uganda passport/visa compliant photo — no manual editing. ## What it does 1. In-app camera (CameraX) with an oval framing guide 2. On capture: on-device ML Kit face detection (landmarks + eyes-open) and selfie segmentation to isolate the person 3. Background is replaced with plain white 4. Head-height and eye-line are measured from the real detected face/hair silhouette and cropped to match the official Uganda spec (51x51mm, 600x600px @ 300dpi) 5. Basic quality checks: no face / multiple faces, eyes closed, too dark, too bright, too blurry, face too close to frame edge 6. On confirm, the photo is saved straight to the device's Gallery via MediaStore (album: **Autopass Uganda**) — no manual export step All processing is on-device and offline (no cloud API, no billing account needed). ## Getting the APK Every push to `main` triggers the **Build Debug APK** GitHub Actions workflow. Once it finishes (Actions tab → latest run), download the `autopass-uganda-debug-apk` artifact, unzip it, and install `app-debug.apk` on your phone (enable "install unknown apps" for your file manager/browser first). This is a debug build, so no signing keystore or secrets are required. ## Project structure ``` app/src/main/java/com/autopass/uganda/ standards/ Uganda passport photo spec constants processing/ Face detection + segmentation + crop math + quality checks camera/ CameraX screen, permissions, framing guide overlay ui/ Result screen, theme util/ MediaStore save MainActivity.kt Screen state machine (Camera → Processing → Result) ```