IMAGE RESTORATION

OUTLINE:
 What is Image Restoration?
 Degraded Images
 Degradation Model
 Common Noise Model
 Restoration Techniques
 Restoration Applications
 What is Image Restoration?
 Image Restoration is a process that attempts to reconstruct
or recover an image that has been degraded by using a
priori knowledge of the degradation phenomenon.
 Image Restoration compensates for data errors, noise and
geometric distortions introduced during the scanning,
recording, and playback operations.
 Image restoration attempts to restore images that have
been degraded.
 Restoration involves modeling of degradation and applying
the inverse process in order to recover the original image.
 Degraded Images
 What caused the image to blur?
 Can we improve the image, or “undo” the effects?
Image Restoration vs Image Enhancement
 Image enhancement purpose is to “improve” an image subjectively while Image restoration
is to remove distortion from image in order to go back to the “original” image (objective
process).
 The concerns of the image restoration are the removal or reduction of degradations which
are included during the acquisition of images e.g.; Noise, pixel value errors, out of focus
blurring or camera motion blurring using prior knowledge of the degradation phenomenon.
This means it deals with the modelling of the degradation and applying the process (inverse)
to reconstruct the image. The image restoration has got a wide scope of usage.
 Degradation Model

Spatial Domain:

Frequency Domain :

Matrix = G = HF + N
 Common Noise Models
 Gaussian
- Most common model
 Rayleigh
 Erlang or Gamma
 Exponential
 Uniform
 Impulse
- Salt and pepper noise
Noise Effects on Histogram
Estimation of Degradation Model
 Whether the spatial or frequency domain or Matrix, in all
cases knowledge of degradation function (H) is important.
 Three ways to estimate the H:
- By Observation - No knowledge of degraded function is given, try
to estimate the degraded function in the region which has simpler
structure.
- By Experimentation – try imaging set up similar to original (impulse
response and impulse simulation).
- Mathematical Modeling – use mathematical models.
 After approximating the degradation function, apply the
BLIND CONVOLUTION to restore the original image.
 Restoration Techniques
 Median Filter
 Adaptive Filter
 Linear Filters
 IBD (Iterative Blind De-convolution) method
 Non Negative and Support Constraints Recursive Inverse Filtering (NAS-RIF)
 Super Resolution Restoration Algorithm based on Gradient Adaptive Interpolation
 Deconvolution Using a Sparse Prior
 Block Matching
 Wiener Filter
 Deconvolution using Regularized Filter (DRF)
 Lucy-Richardson Algorithm Techniques
 Restoration Applications
 Astronomical applications characterized by Poisson noise,
Gaussian noise;
 Medical imaging such as computerised tomography(CT) and
magnetic resonance imaging (MRI)
 Over the multispectral bands of satellite imagery , multispectral
image restoration can be carried out in order to improve the
resolution of the captured satellite images
 Improvement of video resolution, the motion blur estimation
can be performed in the real time video image processing
applications.
- END -