EC1009 - Digital Image processing
Part A
1. What is false contouring?
In an image, preserving the spatial resolution constant and reducing the number of
gray levels produces a fine ridge like structures in the smooth areas. This effect is caused
by the use of insufficient gray levels is called as false contouring.
2. State 1 D Hadamard transforms.
The Walsh transform is defined by
Where --------------1-D forward Hadamard Kernel
3. What is the principle of sharpening filters?
The principle objective of sharpening is to highlight the fine details in an image or
to enhance detail that has been blurred either in error or in image acquistition,
4. What is contrast stretching?
Contrast stretching reduces an image of higher contrast than the original by
darkening the levels below m and brightening the levels above m in the image.
5. List the advantages of interactive restoration.
The observer can control the restoration process by tuning the available
Parameters and able to obtain a final result that may be quite adequate for a special
purpose.
6. Why geometric transformation is called rubber sheet transformation?
Geometric transformation is applied to a image to modify the spatial relationship
between the pixels and to restore the image. It is called as rubber sheet transformation
because it may be viewed as printing the image on a sheet of rubber and then stretching
this sheet to some predefined set of rules.
7. What is inter frame redundancy?
It is directly related to the inter pixel correlations within a image. Because the
value of any given pixel can be predicted from the values of its neighbor, the visual
information carried by the individual pixel is relatively small.
8. Write the equation for the fidelity criteria for an image?
A closely related objective fidelity criterion is the mean square signal to the noise
ratio of the compressed - decompressed image.
9. List the problems associated with region growing.
Descriptors alone for region growing can yield misleading results if connectivity
or adjacency information is not used.Another problem with the region growing is the formulation of stopping rule.
10 How normalization can be done for chain code?
The chain code can be normalized with respect to the starting point is that a chain
code can be represented as a circular sequence of direction numbers and redefine the
starting point so that resulting sequence of numbers forms an integer of minimum
magnitude.
Part B
11. a) Describe the convolution property of 2D Fourier transform applicable for images.
Convolution theorem
If f(x) has a fourier transform F(u) and g(x) has a Fourier transform G(u) then
f(x)*g(x) has a fourier transform F(u).G(u).
Convolution in x domain can be obtained by taking the inverse Fourier transform
of the product F(u).G(u).
Convolution in frequency domain reduces the multiplication in the x domain
F(x).g(x) = F(u)* G(u)
These 2 results are referred to the convolution theorem.
b) Discuss in detail about the discrete Walsh transform of an image where N=4.
The Walsh transform is defined by
Where --------------1-D forward Walsh Kernel
The 1-D inverse Walsh transform is expressed as
Where --------------1-D Inverse Walsh Kernel
For N=4
| x u | 0 | 1 | 2 | 3 |
| 0 | + | + | + | + |
| 1 | + | + | + | + |
| 2 | + | + | - | - |
| 3 | + | + | - | - |
12. a) Explain about histogram equalization.
Histogram equalization
12.b.Discuss in detail about sharpening filters.
Sharpening filters in the spatial domain
· Basics
· Sharpening filters
Sharpening filters in the frequency domain.
· Smoothing filters
· Sharpening filters
13. a) Explain in detail image degradation model /restoration process in detail.
· Image degradation model /restoration process diagram
· Degradation model for Continuous function
· Degradation
b) Explain in detail about least mean square filtering with one application
Least mean square filtering
Properties of SVD
· The SVD transform varies drastically from image to image.
· The SVD transform gives best energy packing efficiency for any given
image.
· The SVD transform is useful in the design of filters finding least
Square, minimum solution of linear equation and finding rank of large
matrices.
14. a) Discuss in detail about lossless predictive coding.
Explanation with block diagram
· Reducing interpixel redundancies of closely spaced pixels.
· When predictive coding is used, the image has no need to be decomposed into bit planes or binary images
· New information=Actual value of pixel – Predicted value of that pixel
· The lossless predictive coding model includes both an encoder and a decoder
· The basic components of a predictive encoder are predictor, nearest integer block and symbol encoder
· The components of the predictive decoder are symbol decoder and predictor
14. b) Discuss in detail about various image compression standards.
· Binary image compression standards
· JPEG Standard(Continuous tone still Image compression standards)
· MPEG Standard(Video compression standards)
15. a) Explain in detail about Fourier descriptors and its usage
Boundary descriptors
. Simple descriptors.
. Fourier descriptors
b) i) Region growing by pixel growing.
Region growing is the process of grouping pixels or sub regions into larger regions based on a predefined similarity criteria
· Selection of seed points
· Selection of similarity criteria
· Formulation of a stopping rule
· Region growing algorithm
ii) Region split and merging techniques.
Region split and merging is a segmentation process in which an image is initially subdivided into a set of arbitrary, disjoint regions and then the regions are merged and/or split to satisfy the basic conditions
Quadtree;
A quadtree is a tree in which the nodes have exactly four descendants. The splitting technique can be represented using this quadtree
No comments:
Post a Comment