README.md

metadata

title: Smart Object Size Estimator
emoji: 📏
colorFrom: blue
colorTo: green
sdk: gradio
sdk_version: 4.44.0
app_file: app.py
pinned: false
license: mit

📏 Smart Object Size Estimator

An AI-powered web application that estimates real-world sizes of objects in images using depth estimation and segmentation.

🚀 Features

• AI-Powered Detection: Uses Depth Anything V2 for depth estimation and SAM for object segmentation
• Reference Object System: Click on any detected object and specify its known size to calibrate measurements
• Camera Presets: Pre-configured settings for common smartphones (iPhone, Samsung, Google Pixel)
• Real-time Processing: Interactive interface with immediate feedback
• Robust Scaling: Works with images of any size and format

🛠️ How to Use

1. Upload an Image: Choose any photo containing objects you want to measure
2. Select Camera Type: Pick your device from presets or use custom focal length values
3. Process Image: Click "Process Image" to detect objects (they'll be numbered and colored)
4. Set Reference: Select an object you know the size of and enter its dimensions
5. Calculate Sizes: Get measurements of all detected objects!

📷 Camera Settings

The app includes presets for:

• iPhone 12/13/14 (Main & Ultra Wide cameras)
• Samsung Galaxy S21/S22
• Google Pixel 6/7
• Generic Smartphone
• Custom (manual focal length input)

�� Tips for Best Results

• Use good lighting and avoid harsh shadows
• Ensure objects are clearly visible and well-separated
• Choose a reference object with a known precise dimension
• Take photos perpendicular to objects when possible
• For best accuracy, use camera-specific presets

🔧 Technical Details

• Depth Model: Depth Anything V2 (ViT-Small)
• Segmentation: Segment Anything Model (SAM)
• Interface: Gradio web interface
• Processing: Automatic image scaling and robust preprocessing

📚 Dependencies

All required packages are listed in requirements.txt. Key dependencies:

• Gradio for web interface
• PyTorch for model inference
• OpenCV for image processing
• Matplotlib for visualizations

🚀 Deployment

This app is designed to run on Hugging Face Spaces. Simply:

1. Upload all files to a new HF Space
2. Ensure the model weights are properly downloaded
3. The app will auto-launch with Gradio

🎯 Use Cases

• E-commerce: Product dimension verification
• Interior Design: Furniture and room measurements
• Manufacturing: Quality control and inspection
• Education: Physics and measurement exercises
• Daily Life: Measuring objects without a ruler

⚠️ Limitations

• Accuracy depends on camera calibration
• Works best with well-lit, clear images
• Small or very thin objects may not segment well
• Depth estimation quality affects final measurements