Update README.md

README.md

@@ -1,435 +1,116 @@
-# 🚀 SPaCial AI - Complete OCR System
-
-> **AR-Style Live Camera Scanner + Telegram Bot + Training Data Collection**
-
-Everything runs on your Hugging Face Space - Deploy in 5 minutes! ⚡
-
----
-
-## 🎯 What This Is
-
-A **complete crowdsourced OCR system** for blueprints and technical drawings:
-
-- 📸 **Live Camera Scanner** - AR-style overlay with real-time detection
-- 🤖 **Telegram Bot** - Easy user interface
-- 📊 **Training Data Collection** - Automatic model improvement
-- ☁️ **Cloud Hosted** - Free on Hugging Face Spaces
-
----
-
-## ✨ Features
-
-### For Users:
-- ✅ Send photo → Get OCR results instantly
-- ✅ Open camera → Point at text → See green boxes live
-- ✅ Validate or correct detections
-- ✅ Download results as JSON
-- ✅ Works on any device (mobile/desktop)
-
-### For You:
-- ✅ Collect training data automatically
-- ✅ Store everything in Telegram (free!)
-- ✅ No database needed
-- ✅ Track improvements over time
-- ✅ Crowdsource corrections
-
----
-
-## 📱 User Experience
-
-### 1. Start Bot
-```
-User → Telegram → @your_bot → /start
-
-Bot shows:
-┌────────────────────────┐
-│ [📸 Live Camera Scan]  │ ← Opens mini app
-│ [📁 Upload Photo]      │ ← Traditional
-└────────────────────────┘
-```
-
-### 2. Live Camera (NEW!)
-```
-Click "Live Camera Scan"
-    ↓
-┌──────────────────────────┐
-│ 📊 Scanning...           │
-│ 5 zones detected         │
-├──────────────────────────┤
-│   🎥 CAMERA VIEW         │
-│                          │
-│   ┌───────┐ 12.5 ± 0.1  │ ← Green boxes
-│   │       │              │   appear live!
-│   └───────┘              │
-│                          │
-│      ┌────┐ 10.0        │
-│      │    │              │
-│      └────┘              │
-├──────────────────────────┤
-│ [Stop]      [Capture]    │
-└──────────────────────────┘
-```
-
-### 3. Upload Photo
-```
-Send photo → Wait 3s → Get results
-
-📊 OCR Results
-✅ Found 15 zones:
-1. 12.5 ± 0.1 (98%)
-2. 10.0 (95%)
-3. 25.3 ± 0.2 (97%)
-...
-
-[✅ All Correct] [✏️ Fix] [📥 JSON]
-```
-
----
-
-## 🏗️ Architecture
-
-```
-┌─────────────────────────────────────────────┐
-│           Hugging Face Space                │
-│                                             │
-│  ┌─────────────────────────────────────┐   │
-│  │         app.py (Backend)            │   │
-│  │  • FastAPI OCR endpoints            │   │
-│  │  • PaddleOCR processing             │   │
-│  │  • Telegram bot handlers            │   │
-│  │  • Webhook receiver                 │   │
-│  │  • Static file serving              │   │
-│  └─────────────────────────────────────┘   │
-│                                             │
-│  ┌─────────────────────────────────────┐   │
-│  │   static/camera.html (Mini App)     │   │
-│  │  • Real-time camera access          │   │
-│  │  • Canvas overlay drawing           │   │
-│  │  • Live OCR visualization           │   │
-│  │  • Telegram Web App integration     │   │
-│  └─────────────────────────────────────┘   │
-└─────────────────────────────────────────────┘
-                    ↕
-┌─────────────────────────────────────────────┐
-│            Telegram Bot API                 │
-│  • Webhook messages                         │
-│  • Photos & commands                        │
-│  • Callback buttons                         │
-│  • Mini app launcher                        │
-└─────────────────────────────────────────────┘
-                    ↕
-┌─────────────────────────────────────────────┐
-│              Users                          │
-│  • Send /start → Get camera button          │
-│  • Click button → Mini app opens            │
-│  • Camera scans → Live boxes appear         │
-│  • Or upload photo → Get results            │
-└─────────────────────────────────────────────┘
-                    ↕
-┌─────────────────────────────────────────────┐
-│       Telegram Channel (Storage)            │
-│  • Training data (JSON)                     │
-│  • Original images                          │
-│  • User corrections                         │
-│  • Free, persistent, backed up              │
-└─────────────────────────────────────────────┘
-```
-
----
-
-## 📦 What's Included
-
-### Backend (`app.py` - 1200 lines)
-- ✅ OCR API endpoints
-- ✅ Telegram bot webhook handler
-- ✅ Training data collection
-- ✅ Photo processing (upload & Telegram)
-- ✅ Interactive buttons (callbacks)
-- ✅ Commands (/start, /help, /status)
-- ✅ Mini app serving
-
-### Mini App (`static/camera.html` - 270 lines)
-- ✅ Camera access (front/back)
-- ✅ Real-time frame capture
-- ✅ OCR API integration
-- ✅ Canvas overlay system
-- ✅ Green boxes + text labels
-- ✅ Status display
-- ✅ Start/Stop/Capture controls
-
-### Documentation (5 files)
-- ✅ `README.md` - This file (overview)
-- ✅ `QUICK_START.md` - 5-minute deploy guide
-- ✅ `TELEGRAM_SETUP.md` - Bot configuration
-- ✅ `MINI_APP_SETUP.md` - Camera app details
-- ✅ `CHANGES_SUMMARY.md` - Technical changes
-
----
-
-## ⚡ Quick Start
-
-### 1. Create Telegram Bot (2 min)
-```
-1. Message @BotFather
-2. /newbot → Name it
-3. Copy token
-```
-
-### 2. Create Telegram Channel (1 min)
-```
-1. Create private channel
-2. Add bot as admin
-3. Get chat ID
-```
-
-### 3. Deploy to HF Spaces (2 min)
-```bash
-git push
-
-# Set env vars in HF Space settings:
-TELEGRAM_BOT_TOKEN=your_token
-TELEGRAM_CHAT_ID=your_chat_id
-SPACE_HOST=your-space-name.hf.space
-```
-
-### 4. Configure Webhook (30 sec)
-```
-Visit: https://your-space.hf.space/telegram/set-webhook?webhook_url=https://your-space.hf.space/telegram/webhook
-```
-
-### 5. Test!
-```
-Telegram → Bot → /start → Click camera button → DONE! 🎉
-```
-
-**[Full instructions in QUICK_START.md](QUICK_START.md)**
-
----
-
-## 🎓 How It Works
-
-### Live Camera Scanning:
-
-```javascript
-// Mini app (camera.html)
-1. Access device camera
-   navigator.mediaDevices.getUserMedia()
-
-2. Capture frame every 2 seconds
-   canvas.toBlob() → image data
-
-3. Send to backend
-   POST /ocr/process
-
-4. Receive zones
-   { zones: [{text, x, y, width, height}] }
-
-5. Draw overlay
-   ctx.strokeRect() → green boxes
-   ctx.fillText() → labels
-
-6. Repeat → Live detection!
-```
-
-### Photo Upload:
-
-```python
-# Backend (app.py)
-1. User sends photo to bot
-2. Telegram webhook → /telegram/webhook
-3. Download photo → process_telegram_photo()
-4. PaddleOCR → detect zones
-5. Format results → format_ocr_results_for_telegram()
-6. Send back to user with buttons
-7. User validates → send_correction_to_telegram()
-8. Store in Telegram channel → Training data!
-```
-
 ---
-
-## 🔗 Endpoints
-
-### User-Facing:
-- `GET /` - Health check
-- `GET /camera` - Mini app interface
-- `POST /ocr/process` - Upload & process image
-
-### Bot Integration:
-- `POST /telegram/webhook` - Receive messages
-- `POST /telegram/set-webhook` - Configure webhook
-- `GET /telegram/status` - Check configuration
-
-### Training Data:
-- `POST /corrections/submit` - Submit user corrections
-
-### Documentation:
-- `GET /docs` - Interactive API docs (FastAPI)
-
+title: SPaCial AI OCR Service
+emoji: 🔍
+colorFrom: blue
+colorTo: purple
+sdk: docker
+sdk_version: "4.21.0"
+app_file: app.py
+pinned: false
+license: mit
 ---
 
-## 📊 Stats
+# SPaCial AI OCR Service 🔍
 
-### Code:
-- **Backend:** 1,200 lines (Python/FastAPI)
-- **Mini App:** 270 lines (HTML/JS)
-- **Documentation:** 5 markdown files
+Advanced OCR service using PaddleOCR for dimension detection in technical drawings. Perfect for extracting measurements, tolerances, and text from engineering drawings.
 
-### Features:
-- **3 ways to scan:** Camera, upload, send to bot
-- **2 types of storage:** Telegram channel + temporary files
-- **1 deployment:** Everything on HF Spaces
+## 🚀 Features
 
-### Performance:
-- **OCR:** 2-5 seconds per image
-- **Live scan:** Every 2 seconds
-- **Bot response:** <1 second
-- **Total latency:** ~3 seconds end-to-end
+- **High-Accuracy OCR**: Uses PaddleOCR with multiple detection modes
+- **Dimension Detection**: Automatically identifies measurements and tolerances
+- **Vertical Text Support**: Handles rotated text (90°, 180°, 270°)
+- **Interactive Canvas**: Draw rectangles or click to detect text
+- **Real-time Processing**: Fast and efficient text extraction
+- **Multiple Modes**: Fast, accurate, and hardcore detection modes
 
----
-
-## 🌟 Why This Is Special
-
-### Traditional OCR:
-```
-Upload → Wait → Get results → Done
-(No improvement, no feedback loop)
-```
-
-### This System:
-```
-Scan live → See boxes → Validate → Correct → Train model → Improve → Better results
-(Continuous improvement with every user!)
-```
+## 🎯 How to Use
 
-### Unique Features:
-- ✅ **AR-style overlay** - See detections in real-time
-- ✅ **Crowdsourced training** - Users improve the model
-- ✅ **Telegram storage** - No database needed
-- ✅ **Multi-modal** - Camera, upload, or send
-- ✅ **Mobile-first** - Works on any device
-- ✅ **Zero infrastructure** - Just HF Spaces
+### 1. Upload an Image
+- Click "Choose File" to upload a technical drawing
+- Supported formats: JPG, PNG, WebP
 
----
-
-## 📚 Documentation
+### 2. Auto-Detection
+- The system automatically detects all text zones
+- Each zone shows confidence percentage and rotation angle
 
-- **[Quick Start](QUICK_START.md)** - Deploy in 5 minutes
-- **[Telegram Setup](TELEGRAM_SETUP.md)** - Bot configuration guide
-- **[Mini App Details](MINI_APP_SETUP.md)** - Camera app documentation
-- **[Changes Summary](CHANGES_SUMMARY.md)** - Technical overview
+### 3. Manual Selection
+- **Double-click**: Click on any text to detect it precisely
+- **Draw Rectangle**: Drag to draw a rectangle around text for better detection
 
----
+### 4. Zone Management
+- **Rotate**: Click the rotate button to improve vertical text detection
+- **Fit**: Click thumbnail to re-OCR with better accuracy
+- **Delete**: Remove unwanted zones
 
-## 🔧 Configuration
+## 🔧 API Endpoints
 
-### Required Environment Variables:
-```env
-TELEGRAM_BOT_TOKEN=1234567890:ABC...
-TELEGRAM_CHAT_ID=-1001234567890
-SPACE_HOST=your-space-name.hf.space
+### Process Image
 ```
-
-### Optional (Auto-configured):
-```env
-PADDLE_HOME=/app/.paddlex
-PADDLEOCR_HOME=/app/.paddleocr
+POST /ocr/process
+- file: Image file
+- mode: "fast", "accurate", or "hardcore"
+- rotation: 0, 90, 180, 270 degrees
 ```
 
----
-
-## 🐛 Troubleshooting
-
-### Bot not responding?
-```bash
-curl https://your-space.hf.space/telegram/status
+### Process Center Point
 ```
-
-### Mini app won't open?
+POST /ocr/process-center
+- image: Base64 encoded image
+- center_point: {"x": x, "y": y}
+- rectangle_bounds: {"x1": x1, "y1": y1, "x2": x2, "y2": y2}
+- use_rectangle: true/false
+- rotation: rotation angle
 ```
-Check SPACE_HOST env var
-Try: https://your-space.hf.space/camera
-```
-
-### Camera not working?
-```
-1. Grant permissions
-2. Use HTTPS (required)
-3. Try mobile device
-```
-
-### More help:
-- [MINI_APP_SETUP.md](MINI_APP_SETUP.md#troubleshooting)
-- Check HF Space logs
-- Visit `/docs` for API testing
-
----
-
-## 🎯 Use Cases
 
+## 🛠️ Technical Details
 
+### Detection Modes
+- **Fast**: Quick detection for position finding
+- **Accurate**: Balanced speed and accuracy
+- **Hardcore**: Maximum accuracy with preprocessing
 
+### Vertical Text Handling
+- Automatic detection of vertical text regions
+- 90° rotation for better OCR accuracy
+- Proper coordinate mapping back to original image
 
+### Text Correction
+- Built-in correction rules for common OCR errors
+- Tolerance parsing for engineering measurements
+- Dimension validation and formatting
 
+## 📊 Performance
 
-- ✅ **Blueprint scanning** - Extract dimensions
-- ✅ **Technical drawings** - Detect measurements  
-- ✅ **Quality control** - Verify dimensions
-- ✅ **Documentation** - Digitize old plans
-- ✅ **Field work** - Scan on-site
-- ✅ **Training data** - Collect corrections
+- **Processing Time**: 1-3 seconds per image
+- **Accuracy**: 95%+ for clear technical drawings
+- **Supported Languages**: English (optimized for numbers and dimensions)
+- **Image Size**: Up to 10MB, auto-resized for processing
 
----
+## 🔍 Use Cases
 
+- **Engineering Drawings**: Extract dimensions and tolerances
+- **Technical Documentation**: Digitize handwritten measurements
+- **Quality Control**: Verify measurements on drawings
+- **CAD Integration**: Convert scanned drawings to digital data
 
+## 🚨 Troubleshooting
 
+### Common Issues
+1. **Low Confidence**: Try "hardcore" mode or draw a rectangle around text
+2. **Vertical Text**: Use the rotate button to improve detection
+3. **Split Text**: Draw a rectangle to capture the full text
+4. **Small Text**: Ensure image resolution is sufficient
 
+### Tips for Best Results
+- Use high-contrast images
+- Ensure text is clearly visible
+- For vertical text, use the rotate feature
+- Draw rectangles around complex text regions
 
-## 🚀 Future Enhancements
+## 📝 License
 
-### Possible Additions:
-- Video file processing (scan across large blueprint)
-- Multi-frame stitching (combine angles)
-- Export to Excel/CSV
-- Custom models per user
-- Analytics dashboard
-- Multi-language support
-- Voice feedback
-- AR measurement tools
+This project is part of the SPaCial AI system for technical drawing analysis.
 
 ---
 
-
-
-
-## 🤝 Contributing
-
-This is a complete, deployable system. To customize:
-
-
-
-
-1. **Adjust scan interval:** `camera.html` line 150
-2. **Change UI colors:** `camera.html` CSS section
-3. **Modify OCR settings:** `app.py` line 363-381
-4. **Add features:** Extend endpoints in `app.py`
-
----
-
-## 📄 License
-
-MIT License - Use freely for any purpose!
-
----
-
-
-## 🎉 You're Ready!
-
-Deploy to Hugging Face Spaces and start scanning! 
-
-**Read [QUICK_START.md](QUICK_START.md) for step-by-step instructions.**
-
----
-
-Built with ❤️ for the blueprint scanning community
-
-**Questions?** Check the docs or HF Space logs!
+**Ready to extract text from your technical drawings? Upload an image and start detecting!** 🎯