University of Mumbai 2007 M.Sc Information Technology Digital Image Programming & Speech MSc IT, - Question Paper

M. Sc IT, March 2007
Digital Image Programming & Speech

Con. 1390-07 OT-5840
(3 Hours) [Total Marks: 75

N.B. (1) Attempt any 5 ques. but not more than 3 ques. from any part.
(2) Answer to the 2 parts must be written in separate ans books and should be submitted
separately.
(3) Figures to the right indicate full marks. every ques. is of 15 marks.

part I

1. (A) discuss with suitable illustration the process of digital image acquisition process.
(B) describe Histogram of an Image. State the characteristics of histogram equalization.
(C) Write short note on Image Averaging.

OR

2. (A) define in brief the fundamental steps in Image Processing.
(B) define various kinds of neighborhood of a pixel. discuss how neighborhood of a pixel can be
used to obtain connected components for subset of pixels of image?
(C) discuss Quantization with example.

3. (A) elaborate filters in image processing? discuss in brief Idea Low pass filters.
(B) Compare and contrast Error-free and Lossy compression techniques. provide example of every.
(C) State the properties of Fourier transform.

OR

4. (A) discuss with example LZW coding technique.
(B) What is brightness adoption and discrimination? discuss with suitable illustrations.
(C) describe fidelity criteria and state importance.

5. (A) elaborate Morphological Operations? Briefly define Opening and Closing using suitable
structuring element.
(B) What is thickening? discuss with example.
(C) Write short note on Logical Operation on a Digital Image.

OR

6. (A) discuss Region based segmentation technique.
(B) define with block diagram the model of Lossless Predictive Coding techinique.
(C) describe Hit-or-Miss transform.

part II

7. (A) With the help of suitable model, discuss the sound production mechanism in human being. [5]
(B) define the production of nasal consonant and the nature of the spectrum of nasal consonants. [5]
(C) discuss acoustic phonetic vowel classifier. [5]

OR

8. (A) discuss the role of physics (acoustics) and signal processing in developing the speech recognition
branch. [5]
(B) Briefly discuss the meaning of subsequent knowledge sources: [5]
Acoustic, lexical, syntactic, semantic and pragmatic.
(C) elaborate the issues associate with acoustic-phonetic approach to speech recognition. [5]

9. (A) Draw the diagram of complete bank of filter analysis model. define the effect of every block on speech signal. [8]
(B) Writing the LPC analysis equations, provide the matrix form presentation of autocorrection function and
predictor coefficients. [Derivation of the matrix form is not expected]. What does the subsequent LPC
parameter represent: [7]
N, M, P, Q and K.

OR

10. (A) discuss the importance of speech end point detection. elaborate the different approaches adopted for
speech end point detection. [8]
(B) explain the end point constraint, monotonicity conditions and local continuity constraints and their
effects on time alignment procedure. [7]

11. (A) discuss how 1 can deal with advance conditions in speech recognition. [8]
(B) What is HMM? What is the underlying assumption of this model? define the kinds of HMM. [7]

12. (A) discuss voice repertory dialer and automated call kind recognition. [8]
(B) What is subword speech unit? elaborate the possible options for such units? [7]

Earning:   Approval pending.