IT-703-D Digital Image Processing - Nov 2023 Question Paper

Under each question, How to answer lists the points an answer should cover, outlined by CampusPrep from the syllabus. Check the details against your textbook.

1. 7 Marks
   aDescribe image formation in the eye with brightness adaptation and discrimination.
   How to answer

   • Structure of human eye relevant to imaging: lens, retina, rods/cones, fovea
   • Image formation as optical projection onto retina
   • Brightness adaptation: range, Weber ratio, adaptation levels
   • Brightness discrimination: Weber-Fechner, Mach bands / simultaneous contrast with diagram
2. 7 Marks
   bWhat are the different approaches to estimate the noise parameters in an image? Explain.
   How to answer

   • Sources and models of image noise
   • Periodic vs random noise estimation approaches
   • Estimation from image strips, histograms and test patterns
   • Use of mean, variance and PDF parameters from smooth regions
3. 7 Marks
   aExplain about image sampling and quantization.
   How to answer

   • Concept of sampling: continuous to discrete, sampling grid
   • Quantization: gray-level discretization, bit depth
   • Effects on resolution, aliasing, false contouring
   • Diagram of sampler-quantizer model
4. 7 Marks
   bWith necessary equations, explain the concept of homomorphic filtering.
   How to answer

   • Illumination-reflectance model $f(x,y)=i(x,y)r(x,y)$
   • Log, Fourier transform, filter $H(u,v)$, exp steps
   • Block diagram of homomorphic system
   • Simultaneous dynamic range compression and contrast enhancement
5. 7 Marks
   aExplain about histogram processing of color images.
   How to answer

   • Color histograms in RGB/HSI models
   • Per-channel vs intensity-only equalization/specification
   • Preserving hue/saturation while processing intensity
   • Applications in enhancement and segmentation
6. 7 Marks
   bProve that both the 2-D continuous and discrete Fourier transform are linear operations.
   How to answer

   • Definitions of 2-D continuous and discrete Fourier transform
   • Linearity property: transform of $af+bg = aF+bG$
   • Proof using linearity of integral for continuous case
   • Proof using linearity of summation for discrete case
7. 7 Marks
   aExplain image sharpening using Butterworth highpass and Gaussian highpass filters.
   How to answer

   • Highpass filtering concept for sharpening in frequency domain
   • Butterworth highpass transfer function and order effect
   • Gaussian highpass transfer function and smooth response
   • Comparison of ringing, cutoff $D_0$ and results
8. 7 Marks
   bExplain about image restoration using minimum mean square error filtering.
   How to answer

   • Degradation-restoration model and MMSE criterion
   • Wiener / minimum mean square error filter equation
   • Estimation of degradation $H$ and noise/signal power spectra
   • Advantages and limitations in restoration
9. 7 Marks
   aDiscuss about region based segmentation.
   How to answer

   • Principle of grouping pixels into homogeneous regions
   • Region growing by pixel aggregation with seed points
   • Region splitting and merging / split-and-merge algorithm
   • Advantages, limitations and applications
10. 7 Marks
    bWhat is the need of compression? Explain about vector quantization method.
    How to answer

    • Need for compression: storage, bandwidth, transmission time
    • Principle of vector quantization: codebook and codevectors
    • Encoding-decoding steps and distortion measure
    • Advantages and applications in image compression
11. 7 Marks
    aDifferentiate between region oriented segmentation and motion based segmentation.
    How to answer

    • Basis: homogeneity vs motion cues, static vs video images
    • Methods and features used in each approach
    • Handling of illumination, occlusion and noise
    • Tabular comparison with applications
12. 7 Marks
    bExplain the following morphological operations:

    1. i)Erosion
    2. ii)Dilation
    How to answer

    • Definition of structuring element and set operations
    • Erosion: definition $A \ominus B$, shrinking effect
    • Dilation: definition $A \oplus B$, growing effect
    • Duality and simple examples with diagrams
13. 7 Marks
    aExplain the following morphological algorithms.

    1. i)Boundary extraction
    2. ii)Hole filling
    How to answer

    • Morphological basis using erosion and dilation
    • Boundary extraction formula $\beta(A)=A-(A\ominus B)$
    • Hole filling by conditional dilation algorithm steps
    • Illustrative examples and applications
14. 7 Marks
    bDraw a neat block diagram for encoder and decoder in a lossy predictive coding model and explain its working.
    How to answer

    • Structure of lossy predictive encoder and decoder blocks
    • Neat block diagram with predictor, quantizer, feedback
    • Working: prediction error quantization and reconstruction
    • Effect of quantization on distortion and compression
15. 14 Marks
    Write short notes:

    1. a)Texture analysis
    2. b)Image point operations
    3. c)Wiener filtering
    4. d)Object recognition
    How to answer

    • Texture analysis: statistical, structural, spectral approaches
    • Image point operations: single-pixel mapping functions
    • Wiener filtering: MMSE restoration formula and use
    • Object recognition: features, matching and classification stages