👶 Smart Baby Monitor & RTSP Recorder

AI-Powered Sleep & Safety Monitoring System

Transform your IP camera into an intelligent baby monitoring system with real-time AI detection, sleep tracking, and safety alerts.

🌟 Features

🤖 AI-Powered Detection

Person Detection: Advanced YOLOv8 model for accurate human detection
Smart Tracking: DeepSORT algorithm for consistent person tracking across frames
Child Selection: Click-to-select or automatic smallest-person detection

😴 Sleep & Wake Monitoring

Movement Detection: Tracks child movement with configurable sensitivity
Sleep State Tracking: Automatically detects when child falls asleep (customizable duration)
Wake Alerts: Instant notifications when child wakes up and starts moving
Visual Indicators: Real-time sleep/wake status overlay

🛡️ Safety Features

Fall Risk Detection: Monitors if child moves near bed edges
Safe Zone Definition: Automatic bed detection with customizable safety margins
Real-time Alerts: Immediate notifications for potential safety risks
Persistent Monitoring: Tracks multiple consecutive frames before triggering alerts

📹 Recording & Storage

Automatic Segmentation: Videos split into configurable time segments (default: 30 minutes)
Organized Storage: Files organized by date and time blocks for easy access
Dual Recording Modes: Save raw footage or AI-annotated videos with detection overlays
Snapshot Capability: Instant photo capture with keyboard shortcut

🔄 Smart Connectivity

RTSP Stream Support: Compatible with IP cameras supporting RTSP protocol
Auto-Reconnection: Intelligent reconnection on network interruptions
Buffering Optimization: Low-latency streaming with minimal delay
Connection Status: Visual indicators for stream health

👥 Multi-User Web Interface

Secure Authentication: Login/signup system with role-based access control
Family Management: Relationship-based user profiles (Mother, Father, Guardian, Caregiver)
Stream Permissions: Admin control over who can view the live stream
Real-time Monitoring: See which family members are currently watching
Activity Tracking: Monitor user sessions and login history
Admin Dashboard: Complete user management with enable/disable capabilities

📋 Prerequisites

Python 3.8+ installed on your system
IP Camera with RTSP support (tested with TP-Link Tapo cameras)
Network Connection between your computer and camera

🔧 Installation

Run with Docker (Recommended)

Clone the Repository

git clone https://github.com/codeperfectplus/AI-Baby-Monitor.git
cd AI-Baby-Monitor

Configure environment:

cp .env.example .env
# Edit .env with your RTSP camera credentials

Deploy the application:
```
./deploy.sh
```
Access the Web Interface Open your web browser and go to:
```
http://localhost:8847
```
Default Login Credentials
- Username: admin
- Password: password
- ⚠️ Change password after first login!

Run with Python Script

Clone the Repository

git clone https://github.com/codeperfectplus/AI-Baby-Monitor.git
cd AI-Baby-Monitor

Install Dependencies
```
pip install -r requirements.txt
```

Configure Camera Settings

Copy and edit the environment file:

cp .env.example .env

Edit .env with your camera credentials:

# RTSP Camera Configuration
RTSP_USERNAME=your_camera_username
RTSP_PASSWORD=your_camera_password
RTSP_IP=192.168.1.100
RTSP_PORT=554
RTSP_STREAM=stream1

Run the Web Application
```
python app.py
```
Access the Interface Open your browser and navigate to:
```
http://localhost:8847
```

📷 Camera Setup Guide

🔍 Finding Your Camera's RTSP URL

For TP-Link Tapo Cameras:

Enable RTSP in the Tapo app:
- Open Tapo app → Select your camera
- Go to Settings → Advanced Settings → Camera Account
- Enable "RTSP Authentication" and set username/password

RTSP URL Format:

rtsp://username:password@camera_ip:554/stream1

For Other IP Cameras:

Common RTSP URL patterns:

Hikvision: rtsp://username:password@ip:554/Streaming/Channels/101
Dahua: rtsp://username:password@ip:554/cam/realmonitor?channel=1&subtype=0
Axis: rtsp://username:password@ip:554/axis-media/media.amp
Foscam: rtsp://username:password@ip:554/videoMain

🌐 Network Configuration

Find Camera IP Address:

# Scan your network for connected devices
nmap -sn 192.168.1.0/24

# Or check your router's admin panel

Test RTSP Connection:

# Using VLC Media Player
vlc rtsp://username:password@camera_ip:554/stream1

# Using FFmpeg
ffprobe rtsp://username:password@camera_ip:554/stream1

👥 Web Interface & User Management

🔐 Authentication System

The application provides a comprehensive web-based interface with secure user management:

User Roles & Relationships

Father - Full administrative access
Mother - Full administrative access
Guardian - Limited monitoring access
Caregiver - Supervised access with restricted features

Admin Features

User Management: Create, edit, enable/disable user accounts
Stream Control: Grant or revoke streaming permissions per user
Activity Monitoring: See who's currently watching and their login history
Relationship Management: Assign family roles to users

Family Monitoring Dashboard

Active Users Display: Real-time view of family members currently monitoring
Session Tracking: Monitor login times and activity duration
Permission Management: Control access to live stream on per-user basis
Login History: Track family member access patterns

Security Features

First-time Password Change: Mandatory password update on initial login
Session Management: Automatic timeout and secure session handling
IP Tracking: Log access attempts with IP address and browser information
Account Status Control: Enable/disable accounts without deletion

🌐 Web Interface Navigation

Main Dashboard (`/`)

Live RTSP stream with AI detection overlays
Real-time sleep/wake status monitoring
Family activity indicators

User Management (`/auth/user-management`)

Add new family members with relationship assignments
Edit existing user profiles and permissions
View comprehensive login history
Enable/disable streaming access per user

Authentication Pages

Login (/auth/login) - Secure access with remember me option
Signup (/auth/signup) - New user registration with relationship selection
Password Management (/auth/change-password) - Secure password updates

🔴 Real-Time Monitoring Features

👨‍👩‍👧‍👦 Family Activity Tracking

Live User Status: See which family members are currently watching
Session Duration: Track how long each user has been monitoring
Relationship Display: Identify watchers by their family role (Mother, Father, etc.)
Multi-User Support: Multiple family members can monitor simultaneously

🎛️ Admin Control Panel

Stream Access Control: Enable/disable streaming for specific users
User Account Management: Create accounts for family members with specific relationships
Login History: Comprehensive audit trail of family member access
Real-time Permissions: Instantly grant or revoke monitoring access

🔐 Security & Privacy

Relationship-Based Access: Different permission levels based on family relationship
Secure Sessions: Automatic timeout and secure session management
Activity Logging: Track all user interactions for security purposes
Password Policies: Enforce strong passwords and mandatory password changes

🎮 Controls & Interactions

⌨️ Keyboard Shortcuts

Key	Action	Description
`Q`	Quit	Exit the application
`P`	Pause/Resume	Toggle recording pause
`S`	Snapshot	Capture instant photo
`R`	Raw/Annotated	Switch between raw and AI-annotated recording
`C`	Clear Selection	Reset child selection and sleep detection
`Z`	Sleep Detection	Toggle sleep/wake monitoring

🖱️ Mouse Interactions

Left Click: Select child to track (click on person in video)
Auto-Selection: System automatically tracks smallest person if no manual selection

📱 Real-time Notifications

The system sends desktop notifications for:

✅ Child selection confirmation
😴 Sleep state detection
🚨 Wake-up alerts
⚠️ Fall risk warnings

📁 File Organization

~/baby-monitor-recordings/
├── 2025-08-13/           # Date folder
│   ├── 00-06/            # Time block (midnight to 6 AM)
│   │   ├── tapo_023015.avi
│   │   └── tapo_053015.avi
│   ├── 06-12/            # Morning block (6 AM to noon)
│   └── 12-18/            # Afternoon block (noon to 6 PM)
└── 2025-08-14/
    └── ...

~/logs/
└── detections.log        # All alerts and events log

⚙️ Configuration Options

🎯 Detection Settings

Located in config/settings.py:

CONFIDENCE_THRESHOLD = 0.4    # AI detection confidence (0.0-1.0)
TARGET_FPS = 15.0            # Video frame rate (lower = less CPU usage)

😴 Sleep Detection

MOVEMENT_THRESHOLD = 30       # Pixels of movement to detect activity
SLEEP_TIME_SEC = 150         # Seconds of stillness = sleep (2.5 minutes)
WAKE_NOTIFICATION_COOLDOWN = 60  # Seconds between wake notifications

🛡️ Safety Monitoring

SAFE_MARGIN_RATIO = 0.15     # Safety zone margin (15% inside bed)
RISK_FRAMES_THRESHOLD = 10    # Frames outside safe zone before alert
ALERT_COOLDOWN_SEC = 20      # Seconds between repeated alerts

📹 Recording Options

SEGMENT_MINUTES = 30         # Minutes per video file
TIME_BLOCK_HOURS = 6         # Hours per folder block
SAVE_ANNOTATED = True        # Save AI-annotated videos
SHOW_PREVIEW = True          # Show live preview window

🔧 Troubleshooting

❌ Connection Issues

Problem: "Cannot open RTSP stream" Solutions:

Verify camera IP address: ping your_camera_ip
Check username/password in .env file
Confirm RTSP is enabled on camera
Test with VLC: vlc rtsp://user:pass@ip:554/stream1

Problem: "Connection lost frequently" Solutions:

Check WiFi signal strength
Reduce video quality on camera
Use wired connection if possible

🖥️ Performance Issues

Problem: High CPU usage Solutions:

Lower TARGET_FPS in config (try 10 or 5)
Reduce camera resolution
Set SHOW_PREVIEW = False for headless mode

Problem: Video files corrupted Solutions:

Check disk space availability
Ensure stable power supply
Try different video codec in recording module

🤖 Detection Issues

Problem: Child not detected Solutions:

Lower CONFIDENCE_THRESHOLD in config (try 0.3)
Ensure good lighting
Check if child is clearly visible
Manually click on child to select

🔧 Module Issues

Problem: Import errors in modular version Solutions:

Run from project root directory
Test modules: python dev.py test
Check installation: python dev.py status

🎨 Visual Interface

📺 Live Preview Window

The application displays a rich, real-time interface showing:

🟢 Green boxes: Selected child being tracked
🔵 Blue boxes: Other detected persons
🟡 Yellow boxes: Detected beds and furniture
🟢 Green rectangle: Safe zone boundary
💤 Sleep indicators: "SLEEPING [ZZZ]" overlay
⚠️ Risk alerts: "RISK: Near edge!" warnings
🎯 Child info: Track ID, confidence, and class
📊 Status display: Sleep/wake state with color coding

🎛️ Status Overlays

Top Right: Child tracking information and confidence scores
Center: Large wake-up alerts with attention-grabbing colors
Bottom: Interactive instructions and keyboard shortcuts
Edges: Safety zone visualization and risk warnings

📊 Technical Specifications

🧠 AI Models

Object Detection: YOLOv8n (optimized for CPU)
Tracking Algorithm: DeepSORT with cosine distance matching
Processing Mode: CPU-only (CUDA disabled for stability)

🎥 Video Processing

Codecs: MJPG (primary), XVID (fallback)
Container: AVI format
Streaming: RTSP over UDP (low latency)
Buffering: Minimal buffering for real-time processing

💾 System Requirements

RAM: Minimum 4GB (8GB recommended)
Storage: ~1GB per hour of recording
CPU: Multi-core processor (Intel i5/AMD Ryzen 5 or better)
Network: Stable connection to camera (wired preferred)

🤝 Contributing

We welcome contributions! Please feel free to:

🐛 Report bugs and issues
💡 Suggest new features
📝 Improve documentation
🔧 Submit pull requests

📋 Development Setup

# Fork the repository
git clone https://github.com/your-fork/AI-Baby-Monitor.git
cd AI-Baby-Monitor

# Create virtual environment
python -m venv venv
source venv/bin/activate  # Linux/Mac
# or
venv\Scripts\activate     # Windows

# Install development dependencies
pip install -r requirements.txt

# Set up environment variables
cp .env.example .env
# Edit .env with your RTSP camera details

# Initialize database and run
python app.py

Testing Authentication Without RTSP

For testing user management features without a camera:

export AUTH_ONLY_MODE=true
python app.py

🏗️ Architecture Overview

The application has evolved into a full-stack web-based monitoring system with Flask backend and real-time capabilities:

📁 Project Structure

AI-Baby-Monitor/
├── app.py                  # 🚀 Flask web application entry point
├── config/                 # ⚙️ Configuration management
│   └── settings.py        # Application settings and RTSP config
├── models/                 # �️ Database models
│   ├── auth.py            # User authentication and login logs
│   └── notification.py    # Notification system
├── api/                    # 🌐 REST API endpoints
│   ├── auth_route.py      # Authentication routes
│   ├── monitor_route.py   # Monitoring endpoints
│   ├── active_users_route.py # Real-time user tracking
│   └── websocket_handlers.py # WebSocket for live streaming
├── services/               # 🔧 Core business logic
│   ├── detection/         # 🤖 YOLO object detection
│   ├── tracking/          # 📍 DeepSORT tracking
│   ├── streaming/         # 📡 RTSP streaming & web sockets
│   ├── monitoring/        # 👶 Sleep & safety monitoring
│   └── recording/         # 🎥 Video recording
├── templates/              # 🎨 HTML templates
│   ├── index.html         # Main dashboard
│   └── auth/              # Authentication pages
├── forms/                  # 📝 WTForms for user input
└── utils/                  # 🛠️ Utility functions

🔄 System Flow

Browser ←→ Flask App ←→ WebSocket ←→ RTSP Stream
    ↓         ↓                        ↓
 Auth DB ← User Mgmt              AI Detection
    ↓         ↓                        ↓
Login Logs  Activity Track        Recording

🌐 Technology Stack

Backend: Flask + Flask-SocketIO for real-time communication
Database: SQLite with SQLAlchemy ORM
Authentication: Flask-Login with session management
Frontend: HTML5, JavaScript, WebSocket for live streaming
AI Processing: YOLOv8 + OpenCV for computer vision
Deployment: Docker + Docker Compose

📝 License

This project is licensed under the MIT License - see the LICENSE file for details.

🙏 Acknowledgments

Ultralytics for YOLOv8 object detection
Deep Sort Realtime for tracking implementation
OpenCV for video processing capabilities
Community contributors for testing and feedback

🌟 Star this repository if it helps you monitor your little ones safely! 🌟

Made with ❤️ by CodePerfectPlus

Name		Name	Last commit message	Last commit date
Latest commit History 24 Commits
api		api
assets/snapshots		assets/snapshots
config		config
forms		forms
models		models
services		services
static		static
templates		templates
utils		utils
.env.example		.env.example
.gitignore		.gitignore
Dockerfile		Dockerfile
LICENSE		LICENSE
README.md		README.md
app.py		app.py
deploy.sh		deploy.sh
docker-compose.yml		docker-compose.yml
requirements.txt		requirements.txt

License

codeperfectplus/AI-Baby-Monitor

Folders and files

Latest commit

History

Repository files navigation