🌱 Seed Image Uploader

A lightweight local network image uploader built with Flask + Pillow — optimised for Raspberry Pi and low-power devices. Run the server, upload photos from your phone over Wi-Fi, and every image is automatically downsampled to 128×128 pixels and saved to your configured folder. The HD original is never written to disk.

Includes a standalone companion script (hd_to_lr_test.py) for live webcam → 128×128 capture.

---

✨ Features

• 📐 Auto 128×128 downsampling — server-side resize on every upload; HD original never saved
• 📱 Phone-friendly UI — fully responsive drag & drop interface
• 🔲 Auto QR Code — scan to open the uploader instantly on any device
• 📦 Multi-file upload — up to 20 images at once, 50 MB each
• 📁 Configurable save folder — one-line change in app.py
• 📊 Live stats — session counter, total files, recent uploads list
• 🌐 LAN only — purely local, no internet required
• 🪶 Lightweight — pure Python, minimal dependencies, boots fast on RPi

---

🔄 Upload Pipeline

Phone / Browser
      │
      │  POST /api/upload  (multipart, ≤ 50 MB each)
      ▼
  Flask  ──►  Pillow: Image.resize((128, 128), LANCZOS)
                                    │
                                    ▼
              ~/ardu_ws/seed_images/lr_<name>_<timestamp>.jpg

Only the 128×128 LR JPEG is ever written to disk.

---

🚀 Quick Start

1 — Install dependencies

git clone https://github.com/your-username/Seed_image_uploader.git
cd Seed_image_uploader

python3 -m venv venv
source venv/bin/activate          # Windows: venv\Scripts\activate

pip install -r requirements.txt

2 — Configure save folder

Edit the top of app.py:

UPLOAD_DIR = Path.home() / "ardu_ws" / "seed_images"   # ← your path here
PORT       = 3000
LR_SIZE    = (128, 128)

3 — Run

python app.py

╔══════════════════════════════════════════╗
║       🌱  Seed Image Uploader  🌱        ║
╠══════════════════════════════════════════╣
║  Local:   http://localhost:3000          ║
║  Network: http://192.168.1.42:3000       ║
╚══════════════════════════════════════════╝
📱 Scan the QR code on the webpage to connect from your phone!

---

📱 Using from Your Phone

1. Phone and RPi/laptop must be on the same Wi-Fi
2. Open http://localhost:3000 on the host device
3. Scan the QR code — the upload page opens on your phone
4. Select images → tap Upload → saved as lr_*.jpg instantly

Or type the Network URL shown in the terminal directly into your phone's browser.

---

📁 Project Structure

Seed_image_uploader/
├── app.py                # Flask server — upload, Pillow resize, QR code, file list
├── hd_to_lr_test.py      # Standalone: webcam → 128×128 → saves to Desktop2
├── requirements.txt      # Python deps: flask, Pillow, qrcode[pil]
├── public/
│   ├── index.html        # Frontend HTML
│   ├── style.css         # Dark glassmorphic UI
│   └── app.js            # Drag & drop, preview, upload, stats
├── .gitignore
└── README.md

---

⚙️ Configuration

Setting	Where	Default
Save folder	app.py → UPLOAD_DIR	~/ardu_ws/seed_images
Output size	app.py → LR_SIZE	(128, 128)
Port	app.py → PORT	3000
Max file size	app.py → MAX_FILE_SIZE_MB	50
Resize filter	app.py → Image.LANCZOS	Lanczos
Allowed formats	app.py → ALLOWED_EXTENSIONS	JPG PNG GIF BMP WEBP HEIC TIFF

---

🌐 API Endpoints

Method	Endpoint	Description
GET	/	Serves the web UI
GET	/api/info	Server IP, port, save path, QR code data URL
POST	/api/upload	Upload images (field: images); returns saved LR filenames
GET	/api/files	List saved LR images with size & timestamp

---

🎥 Standalone Webcam Script

hd_to_lr_test.py captures live webcam frames and saves 128×128 images to two folders simultaneously. Useful for collecting seed images without a phone.

Requires: pip install opencv-python

python3 hd_to_lr_test.py   # press q to stop

Webcam (HD) ──► cv2.resize(128×128, INTER_AREA)
                    ├──► ~/ardu_ws/seed_images/lr_YYYYMMDD_HHMMSS_NNNN.jpg
                    └──► ~/Desktop2/lr_YYYYMMDD_HHMMSS_NNNN.jpg

---

📦 Dependencies

Package	Purpose
Flask	Web framework
Pillow	Image resize to 128×128 JPEG
qrcode[pil]	QR code generation
opencv-python	Webcam script only (optional)

---

🔒 Notes

• Intended for local network use only — add auth before exposing publicly
• HD originals are never stored; Pillow processes them in memory
• Output is always JPEG regardless of input format
• On Raspberry Pi, add to crontab for auto-start on boot:

  @reboot cd /path/to/Seed_image_uploader && venv/bin/python app.py >> ~/uploader.log 2>&1 &
  

---

📄 License

MIT