Question 1:
(a) Describe the steps involved in the performing Image based processing.
(b) Propose a mask using a 3X3 matrix, which would aid in discovering discontinuities, and hence give its suitable response as a mathematical model.
(c) Propose four matrices to be employed as linear masks in line detection.
(d) Give the approach employed by means of Hough Transform to aid in solving the edge-linking problem.
Question 2:
(a) Explain the following Mean Filters employed as Noise Reduction filters:
1. Arithmetic Mean Filter.
2. Geometric Mean Filer.
3. Harmonic Mean Filter.
You are also needed to indicate which type of noise they are best adapted for.
(b) Make a distinction between the Ideal Low Pass and Ideal High Pass filters in the Frequency domain using the Fourier Transform.
(c) In brief give the steps of how you would perform filtering in Frequency Domain.
Question 3:
(a) Median filters do not cater for a dynamic range of pixels in given area, S. Thus to ensure that no loss of image details occur in S, adaptive median filters could be employed. Propose a 2-level algorithm for such a filter.
(b) In Image Restoration, Degradation functions are in general estimated functions. Show how to represent such a function as a Mathematical Model, supposing that the influenced image produced was from an unexpected movement of camera.
Prove also that it can be represented in Frequency Domain as the given expression, G (u,v) = H(u,v) * F(u,v).
(c) Show in a diagram:
1. The relationships between wavelet and scaling functions spaces.
2. A lossy predictive coding model for an encoder in Image Compression paradigm.