42. Optimal unsharp mask - 2011

42. Optimal unsharp mask - 2011 - Purdue University Purdue...

Info iconThis preview shows pages 1–8. Sign up to view the full content.

View Full Document Right Arrow Icon

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: Purdue University Purdue University Research objective and motivations Optimal unsharp mask Modeling/characterization of the imaging pipeline Optimization of the filter parameters Experimental results Purdue University Conventional imaging pipeline for digital image capture. Captured images suffer from the blur and the noise. Want to remove the blur and the noise from the capture. Want an enhancement method with reasonable computatioal complexity for the desktop printing application. Purdue University Conventional unsharp mask (UM). : enhanced image, g [.]: blurred-noisy image, H{.}: highpass filter, M : unshap mask strength. L{.}: lowpass filter. H{.} g[m,n] M HPF Purdue University Advantages of the conventional UM. Linear space-invariant filter. Easily implemented as a spatial domain convolution filter. Computational inexpensive for sharpening operation. Works well with small kernel sizes. Very robust. For different degrees of the blurring, blur PSF shape, etc. Purdue University Problem with the conventional UM Strong sharpening parameter -> ringing in results. Weak sharpening parameter -> insufficient sharpening. Amplification of the background noise -> objectionable in skin areas. 10 20 30 40 50 60 70 80 20 40 60 80 100 120 140 160 180 200 Sharpening with Unsharp Mask (window size=7) original blurred WM filtered Purdue University S. Gullion, P. Baylou and M. Najim (Jun. 1998) Adaptive nonlinear filters for 2D and 3D image enhancement, Signal Processing, Elsevier. Utilized locally varying filter mask. To reduce the noise sensitivity. A separate lowpass filter in direct data path. To achieve the smoothing of the noise. Adaptive UM strength. To prevent the noise amplification and over-sharpening....
View Full Document

Page1 / 33

42. Optimal unsharp mask - 2011 - Purdue University Purdue...

This preview shows document pages 1 - 8. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online