Welcome, Demo User!

How It Works

BlurDetect uses 5 advanced computer vision models to classify blur types:

• Hybrid Laplacian + FFT: Combines spatial and frequency analysis
• Sobel Gradient: Analyzes directional edges
• FFT Spectrum: Frequency domain analysis
• PSF Estimation: Point Spread Function analysis
• CNN Deep Learning: Neural network classification

The primary model (Hybrid Laplacian + FFT) is selected for its balance of accuracy, speed, and reliability without requiring training data.