Vasthu Sense AI Scanner is a full-stack mobile application consisting of an Android app (Kotlin + Jetpack Compose) and a Python Flask backend powered by AI vision models.
1) The user opens the app and is greeted with an animated Welcome Screen.
2) On the Camera Screen, a live camera preview is shown alongside a real-time compass needle (using device sensors) and a direction banner (e.g., "Facing: Northeast → 45°").
3) The user taps the capture button — the app sends the room image along with compass heading and facing direction to the AI backend.
4) The backend processes the image using AI vision (Gemini Vision or GPT-4o-mini), identifies the room type (kitchen, bedroom, living room, etc.), and evaluates it against a comprehensive Vasthu Shastra rules database covering all 8 cardinal directions plus the center (Brahmasthan).
5) The Result Screen displays an overall Vasthu score (0–100) with an animated ring, along with categorized findings:
-> Critical Defects (red) — serious Vasthu violations
-> Warnings (orange) — notable concerns
-> Suggestions (yellow) — improvements to consider
-> Uncertain items (with caution icon) — things the AI couldn't determine with >70% confidence
-> Positive aspects (green) — things already aligned with Vasthu
-> Priority actions — top 3 most important fixes with remedies