Gabor Filter Image Decomposition
Upload an image and watch a bank of Gabor filters decompose it — the same mathematical operation your visual cortex performs every moment you see.
Drop an image here or click to upload
JPG, PNG, or WebP • Max 2 MB
Your Brain Already Does This
🖥️ In Computer Vision
Gabor filters are used in image processing because they're optimal for extracting texture, edge, and frequency information. They're fundamental to face recognition (DeepFace), fingerprint matching (FBI's IAFIS), and texture classification.
The filter bank above — multiple orientations × multiple frequencies — is exactly how computer vision systems analyze an image. Each filter responds to a specific orientation and scale of pattern.
🧠 In Your Visual Cortex
Neurons in your primary visual cortex (V1) have receptive fields that are almost exactly Gabor functions. This was discovered by Hubel & Wiesel (Nobel Prize, 1981) and formalized by Daugman (1985).
Your V1 contains a complete "filter bank" — neurons tuned to different orientations and spatial frequencies, working in parallel. Training these neural filters with Gabor patches improves how efficiently your brain processes visual information, compensating for age-related optical degradation.
How Gabor Filters Work
The Math
G(x,y) = exp(-(x'² + γ²y'²) / (2σ²)) × cos(2π · x'/λ + ψ)
Where:
θ— orientation of the filter (which direction of edges it detects)λ— wavelength of the sinusoidal carrier (1/spatial frequency)σ— standard deviation of the Gaussian envelope (filter size)γ— spatial aspect ratio (elongation of the filter)ψ— phase offset (symmetric vs. anti-symmetric)
📚 Computer Vision Uses
- • Face recognition (DeepFace, OpenCV)
- • Fingerprint matching (FBI IAFIS)
- • Texture segmentation & classification
- • Optical character recognition (OCR)
- • Medical image analysis (retinal scans)
- • Quality inspection (manufacturing)
🔬 Why Gabor Specifically?
Gabor filters achieve the theoretical minimum uncertainty in both spatial and frequency domains (the "uncertainty principle" of signal processing). This makes them optimally efficient — extracting the maximum information with the minimum computational cost. Evolution arrived at the same solution for biological vision.
🧬 Biology = Same Algorithm
Jones & Palmer (1987) showed that simple cell
receptive fields in cat V1 are well-modeled by Gabor functions. The same has been confirmed in
primates. Your brain's "image processing pipeline" uses the same math as OpenCV's cv2.getGaborKernel().