Printed Pages—4 ECS702
(Following Paper ID and Roll No. to be filled in your Answer Book)}
B.Tech.
(SEM. VII) ODD SEMESTER THEORY
EXAMINATION 2013-14
DIGITAL IMAGE PROCESSING
Time : 3 Hours Total Marks : 100
Note :-- Attempt all questions.
1, Attempt any four parts of the following : (5*4=20)
(a) Describe in detail the elements of di igital image processing
system and describe Sampling and Quantization.
(b) Explain the properties of images which can be described
by histogram. Also explain Normalized Histogram.
(c) Explain histogram matching. Perform the histogram
equalization for 8x8 image shown below :
Gray levels} 0 | 1] 2)3 14 15 16] 7
Ny) 4 | 10) 15] 4
of pixels} 9
(d) Explain the 4, 8 and m connectivity of pixels. Explain region,
edge in context with connectivity of pixels.
(e) Explain the need of Histogram Matching (specification).
Deduce the formula for Histogram Matching.
ECS702/DNG-52034 1 [urn Over2.
(f) The following matrix defines a 5x5 image f(x,y). Suppose
smoothing is done to the image using 3x3 neighbourhood in
the spatial domain. Then what will be the new value of
(2,2) using the :
(i) Mean filter
(i) Max filter
(iii) Median filter
(iv) Min filter.
[3 s[6| 4 7
Attempt any four parts of the following : (5x4=20)
(a) Discuss Image smoothing with the following :
(i) Low pass spatial filtering
(i) Median filtering.
(b) Distinguish between spatial domain techniques and
frequency domain techniques of image enhancement.
(c) An image segment is shown below. Let V be the set of
gray level values used to define connectivity in the image.
Compute D4, D8& and Dm distances between pixel
ECS702/DNG-52034 2
pand q for:
@ v= {2,3}
Gi) v= {2,6}.
P;} 2/3) 2] 6] 1
6} 2} 3] 6] 2
5] 3] 2] 3] 5
2/4) 3] 5S] 2
4] 5} 2] 3} 61 q
(d) Consider a 3x3 spatial mask that averages the four closet
neighbours of a point (x,y), but excludes the point itself
from the average.
@) Find the equivalent filter, H(u,v) in the frequency
domain.
(ii) Show that your result is low pass filter.
(e) Find the equivalent filter H(u,v), that implements in the
frequency domain the spatial operation performed by the
laplacian mask.
(f) Prove that 2-D continuous and discrete Fourier transforms
are linear operations.
Attempt any two parts of the following : (10«2=20)
(a) Explain Image degradation/Restoration Process. Explain
all noises with their PDF.
(b) Explain why Band Rejects filters are best suitable for
reducing Periodic noise. Explain all Band Reject filters in
detail. Obtain corresponding expression for Band pass filters.
ECS702/DNG-52034 3 [Turn Over(c) Explain the following :
@ Wiener filter with SNR, MSE ratio for spatial and
frequency domain
Gi) Local noise reduction adaptive filter.
4. Attempt any two parts of the following : (10x2=20)
(a) Explain morphological image processing in context with
set theory. Explain erosion, dilation, opening and closing
with proper example.
(b) Prove the following properties :
@ = (A+By = (Ao BY) and (A 0 BY = (At: BY)
(i) (AcB)oB = AoB and (A-B)-B=A-B
(c) Explain the following Morphological Algorithms :
@ Thinning
Gi) Thickening
(iii) Convex Hull
(iv) Extraction of Connected Components
(v) Region Filling.
5. Attempt any two parts of the following : (10x2=20)
(a) How many degrees of freedom are there in a plane
projective transformation ? Name the properties preserved
under such transformation. Explain Projective and Affine
transformation.
(b) Discuss parametric and non-parametric methods in optimal
thresholding algorithms. Discuss Region Growing
Approach. Also explain split and merge algorithm with
Quadtree. f
(c) Discuss various Edge detectors in detail. What is Image
Registration ? Explain stereo imaging in detail.
ECS702/DNG-52034 4 13425