Running Form Analyzer

An Android app for analyzing running form using ML-powered pose detection.

Features

• Camera & Gallery Support - Take a photo or select from gallery
• Image Cropping - Crop images to focus on the subject
• Automatic Pose Detection - Uses Google ML Kit to detect body pose
• Interactive Skeleton Overlay - Drag joints to fine-tune positions
• Magnifier Loupe - Precision crosshair when adjusting joints
• Segment Locking - Lock body proportions while adjusting pose angles
• Pinch to Zoom - Zoom in for precise adjustments
• Save & Load Analyses - Persist analyses with athlete names for later review
• Search Saved Analyses - Filter saved analyses by athlete name

How It Works

1. Take or select a photo of someone running
2. The app automatically detects the pose and overlays a stick figure
3. Drag any joint to adjust its position
4. Use the lock icon to maintain body segment lengths while rotating joints
5. Use the refresh icon to re-detect the original pose
6. Save the analysis with an athlete name for future reference
7. Access saved analyses from the home screen

Tech Stack

• Flutter - Cross-platform UI framework
• Google ML Kit - On-device pose detection (33 landmarks)
• Riverpod - State management
• Freezed - Immutable data models

Requirements

• Android device (ML Kit pose detection requires Android/iOS)
• Camera permission (for taking photos)
• Storage permission (for gallery access)

Development

# Install dependencies
flutter pub get

# Generate freezed models
dart run build_runner build

# Run on connected device
flutter run

Building

# Build release APK
flutter build apk --release

# APK location: build/app/outputs/flutter-apk/app-release.apk