150 likes | 281 Views
Removing motion blur from a single image. Sources of blur. Object motion. Sources of blur. Object motion Translation of camera. Sources of blur. Object motion Translation of camera Rotation of camera. Sources of blur. Object motion Translation of camera Rotation of camera. Defocus
E N D
Sources of blur • Object motion
Sources of blur • Object motion • Translation of camera
Sources of blur • Object motion • Translation of camera • Rotation of camera
Sources of blur • Object motion • Translation of camera • Rotation of camera • Defocus • Internal camera distortions
PSF = Point Spread Function (PSF) Assume: • Point light source
B = KL + N Convolution model motivation • Assume: • No image plane rotation • No object motion during the exposure • No significant parallax (depth variation) Camera motion is Pure Translation!!! • Violation of assumption:
B = KL + N Convolution model motivation • Assume: • No image plane rotation • No object motion during the exposure • No significant parallax (depth variation) Camera motion is Pure Translation!!! • Experimental validation: 8 subjects handholding DSLR with 1 sec exposure Close-up of dots
Generation rule:B = KL + N Convolution Model • Notations • L: original image • K: the blur kernel (PSF) • N: sensor noise (white) • B: input blurred image +
Fourier Convolution Theorem! How can the image be recovered? Assumptions: • Known kernel (PSF) • Constant kernelfor the whole image • No noise Goal: • Recover Ls.t.: B = KL
De-blur using Convolution Theorem Convolution Theorem:
Recovered Example: Blurred Image PSF
Deconvolution is unstable Example: Noisy case:
FT of original signal Original signal Convolved signals w/w noise FT of convolved signals Reconstructed FT of the signal 1D example: Regularization is required
51 151 191 Regularizing by window Window size: