Open Source Android App for Guide Thermal Cameras with AI Object Detection
A powerful alternative to the official Guide app, supporting multiple Guide Sensmart thermal camera models including the Guide TE211M Thermal Monocular.
- 🎥 Live Thermal Video Stream via RTSP (LibVLC)
- 🤖 YOLO Object Detection with TensorFlow Lite
- 🎨 Multiple Palettes: Whitehot, Blackhot, Bluehot, Greenhot, Ironred, and more
- 📸 Screenshot & Video Recording with gallery storage
- 📁 Integrated Gallery for captured media
- 📶 Auto-WiFi Connect to camera
- ⚡ Real-time Overlay with bounding boxes and labels
- 🌙 Dark Theme optimized for night use
- 🌡️ Emissivity Control (0.01-1.0) with material presets
- 📏 Distance Compensation for accurate temperature readings
- 💧 Humidity Settings for atmospheric correction
- ♨️ Reflected Temperature adjustment
- 🔄 Shutter/NUC Calibration button
- 📱 Automatic Camera Detection - identifies model and enables features
- 🔌 REST API Integration for professional thermal settings
- 📊 System Log Viewer - Real-time diagnostics with color-coded entries
- ⚙️ Advanced Camera Settings - Brightness, Contrast, Image Enhancement
- 🎵 Audio Control - Enable/disable camera microphone
- 📄 Multi-Page Settings - Organized settings with navigation
- 💎 Feature Bounties - Support development via in-app purchases
- 🎉 What's New Dialog - Version update announcements
- 👋 Welcome Screen - First-run introduction for new users
| Series | Models | Features |
|---|---|---|
| TE-Series | TE211M, TE-Mini | Radiometry ✓ |
| C-Series | C640, C400, C800 | Focus ✓, GPS ✓, Radiometry ✓, Audio ✓ |
| D-Series | D400, D384, D192, D160 | Focus ✓, Radiometry ✓ |
| B-Series | B320, B256, B160 | Radiometry ✓ |
| PS-Series | PS600, PS400 | Radiometry ✓ |
Note: Features are automatically detected based on the connected camera model.
- Minimum: Android 7.0 (API 24)
- Target: Android 15 (API 35)
- Architecture: arm64-v8a
The app automatically connects to the camera via WiFi:
| Parameter | Default Value | Description |
|---|---|---|
| SSID | TE Mini-XXXX |
Camera's WiFi hotspot |
| Password | 12345678 |
Default password for all Guide cameras |
| IP | 192.168.42.1 |
Fixed IP of camera in hotspot mode |
| RTSP | rtsp://192.168.42.1:8554/video |
Video stream URL |
| HTTP API | http://192.168.42.1 |
Camera control |
Note: The IP can be changed in Settings if your camera uses a different address.
- Download the latest APK from Releases
- Install on your Android device
- Allow installation from unknown sources if prompted
Prerequisites:
- JDK 17 or higher
- Android SDK (API 35)
- Git
# Clone repository
git clone https://github.com/nacl-dev/NoxVision.git
cd NoxVision
# Set SDK path (if not auto-detected)
echo "sdk.dir=/path/to/Android/Sdk" > local.properties
# Build debug APK
./gradlew assembleDebug
# APK location:
# app/build/outputs/apk/debug/NoxVision-v1.3-debug.apk- Turn on camera and enable WiFi hotspot
- Launch app - automatically connects to camera
- Live stream displays with object detection
- Controls:
- 📷 Take screenshot
- 🎬 Record video
- 🎨 Change palette
- 📁 Open gallery
- Open Settings ⚙️
- Navigate to "Thermische Einstellungen"
- Adjust:
- Emissivity - Match the material you're measuring
- Distance - Set the distance to your target
- Humidity - Current ambient humidity
- Reflected Temperature - Ambient temperature
- Tap "An Kamera senden" to apply
Monitor app activity and troubleshoot issues:
- Open Settings ⚙️
- Tap "System Log"
- View real-time logs with color-coded entries (Info, Success, Error)
The app uses a trained TensorFlow Lite model for thermal detection:
| File | Description |
|---|---|
detect.tflite |
TFLite model (YOLO-based) |
labelmap.txt |
Class labels (person, vehicle, etc.) |
The model is specifically trained for thermal images and detects people and vehicles in thermal imagery.
app/src/main/
├── java/com/noxvision/app/
│ ├── MainActivity.kt # Main app entry point
│ ├── CameraApiClient.kt # REST API client
│ ├── CameraSettings.kt # Settings persistence
│ ├── DeviceInfo.kt # Camera model detection
│ ├── WhatsNewData.kt # Version changelog
│ ├── ui/
│ │ ├── VideoStreamScreen.kt # Main livestream UI
│ │ ├── SettingsScreen.kt # Multi-page settings
│ │ ├── ThermalSettingsScreen.kt # Thermal measurement UI
│ │ ├── SettingsComponents.kt # Reusable components
│ │ ├── theme/ # Compose Theme (Material 3)
│ │ ├── components/ # Custom UI components
│ │ └── dialogs/ # Dialog screens
│ │ ├── LogDialog.kt # System log viewer
│ │ ├── AboutDialog.kt # App information
│ │ ├── WelcomeDialog.kt # First-run welcome
│ │ ├── WhatsNewDialog.kt # Version updates
│ │ ├── GalleryDialog.kt # Media browser
│ │ └── PreviewDialog.kt # Media preview
│ ├── billing/ # In-app purchases
│ │ ├── BillingManager.kt
│ │ ├── FeatureBountyRepository.kt
│ │ └── FeatureBountyScreen.kt
│ ├── detection/
│ │ └── ThermalObjectDetector.kt # TFLite YOLO detector
│ ├── network/
│ │ └── WiFiAutoConnect.kt # Auto WiFi connection
│ ├── util/
│ │ ├── AppLogger.kt # Logging utility
│ │ └── MediaUtils.kt # Media capture/download
│ └── data/
│ └── MediaModels.kt # Data classes
├── assets/
│ ├── detect.tflite # AI model (YOLOv8)
│ └── labelmap.txt # Class labels
└── res/
├── drawable/ # Palette images
└── values/ # Strings, Colors
- Language: Kotlin 2.0
- UI: Jetpack Compose
- Video: LibVLC
- AI: TensorFlow Lite
- HTTP: OkHttp / HttpURLConnection
- Images: Coil
# Debug (for testing)
./gradlew assembleDebug
# Release (for distribution)
./gradlew assembleReleaseContributions are welcome! See CONTRIBUTING.md for details.
main- Stable versionbeta- Development & new features
- Settings screen for camera IP
- Support for additional camera models ✅ v1.2
- Thermal measurement settings ✅ v1.2
- Export to video formats
- System logging & diagnostics ✅ v1.3
- Advanced camera settings (brightness, contrast) ✅ v1.3
- Multi-page settings navigation ✅ v1.3
- In-app purchase system ✅ v1.3
- Temperature measurement display on live view
- Improved AI model with higher accuracy
- Optimized Wildlife AI Model - Training for deer, boar, fox, rabbit detection
- Distance Estimation - Approximate distance to detected animals
- Hunt Mode - Minimal UI, reduced brightness, one-hand operation
- Silent Alerts - Vibration feedback instead of sounds
- Quick Record - Hardware button mapping for instant recording
- Battery Optimization - Extended runtime for long sits
- Offline Maps - Mark hunting grounds and wildlife trails
- Moon Phase Display - Hunting conditions at a glance
- Shot Documentation - Log shots with timestamp and location
- Tracking Mode - Support for blood trailing after a shot
- Custom Palettes - User-defined color schemes for better contrast
Q: Why this app instead of the official Guide app?
The official app has performance issues and missing features. This open-source alternative offers better performance, AI detection, and professional thermal settings.
Q: Does the app work with other thermal cameras?
Yes! NoxVision supports the full Guide Sensmart lineup including C-Series, D-Series, B-Series, TE-Series, and PS-Series cameras.
Q: What is emissivity and why does it matter?
Emissivity is how well a surface emits thermal radiation. Different materials have different values (human skin ≈ 0.95, polished metal ≈ 0.3). Setting it correctly improves temperature accuracy.
Q: Do I need Android Studio?
No! You can build the APK with VS Code and Gradle. See Installation above.
MIT License - You can freely use, modify, and distribute the app.
- Guide Sensmart for the hardware
- LibVLC for video streaming
- TensorFlow Lite for AI inference
⭐ If you like this project, give it a star!