Domain Track : Advance Video Processing, Computer Vision and Machine Learning

Teacher

Prof. Dr. Jharna Majumdar

Category

Domain Courses

Course Attendees

Still no participant

Course Reviews

Still no reviews

Domain Track Title : Advance Video Processing, Computer Vision and Machine Learning

Track Total Credits: ABC1234 ( 30 Credit)

Courses Division:

 

  1. Image Processing, Computer Vision, Video Processing -- (2-2-2)
  2. Advanced Video Processing and Its  Applications -- (3-3-0)
  3. Machine Learning and ANN -- (2-2-0)
  4. Embedded Hardware Board -- (2-2-0)
  5. Control System -- (2-2-0)
  6. Project -- (0-0-6)

Domain Track Objectives:

  • The students will be trained in Image Processing with Hands-on, which are required as basics to do further studies in these Advance Areas of Technology.
  • This Multidisciplinary Course is meant for students from the Dept. of CSE, ECE, EEE and MECH.
  • Core companies who are looking for trained manpower to be able to work on varieties of technology areas.
  • Students will be given Hands-on Exercises to understand the Fundamental Concept of each algorithm they learn.

Domain Track Learning outcomes:

  • To take the course forward to a specific Domain entitled “Advance Video Processing, Computer Vision and Machine Learning”.
  • The Multidisciplinary Course will on Exercises. Start with brief fundamentals of Image Processing, Computer Vision and Video Processing with Hands-on.
  • After completion of the Course, there are different areas where student will be given projects like Unmanned Vehicle, Video Analytics, Autonomous Robots and associated technology, Advance Video Processing for real time application in Industrial Product Development, Machine Learning, AI and Deep Neural Network and their applications in Autonomous Systems.

Domain Syllabus:

 

Course - 1: Image Processing, Computer Vision, Video Processing -- CUTM1562(2-2-2)

Module I: Image Processing - Fundamentals                                                          (10 hrs)

Introduction to Digital Image, Representation;

Image Processing to Computer Vision: Low Level, Mid Level and High Level Image Processing with examples.

Sensing and Acquisition, Sampling and Quantization, Image Resolution and Storage, Convolution, Basic Relationship between pixels; Monadic Operators

Module II: Image Histogram, Image Enhancement                                                 (10 hrs)

Histogram: Formation of Image Histogram; Characteristics Properties of Images derived from Histogram.                                                         

Algorithms for Enhancement of Images: Contrast Stretching, Linear, Piecewise; Non-Linear - Logarithmic, Exponential, Gamma Correction.

Histogram Based Enhancement: Histogram Equalization; Histogram Specification.

Module III: Edge Detection Algorithms                                                                   (10 hrs)

Mathematical Foundation: First Order and Second order Derivative.

Edge Models: Step, Ramp and Roof Edge.

Algorithms: Robert, Sobel, Prewitt, Laplacian.

Color Models                                                                                                

Color Image Fundamentals:  Light and Color, Chromatic and Achromatic.

Color Models: RGB, HSI, CMY, CMYK, YIQ, YUV, Color Conversions.

Module IV: Fundamentals of Video Processing                                                     (10 hrs)

Video data, video acquisition, video representation, general mathematical operations for video processing

Sampling in spatial and temporal domains, sampling conversion, video to frames and frames to video

Pre-filter in video cameras, interpolation filter in video displays, Fourier analysis of video sequence, spatial frequency, temporal frequency, temporal frequency caused by motion.

Module V: Video data acquisition and formats                                                     (10hrs)

Video data representation, sensors for video acquisition, working of digital camera (block diagram), camera resolution and color conversion, types of video cameras

Visible, IR, thermal and other forms of video data, difference between medical video and generic video

Color perception and specifications, color representation, video formats (NTSC, SECAM, PAL, VGA, QVGA, SVGA, XGA, HD, WUXGA, and QSXGA), video storage requirements, video indexing, video summarization and retrieval

Practice: Hands-On Exercises

 

Module I to III:

  1. Read and Display an Image File. Perform the following operations on the image and display the output on the screen:
    • Transpose
    • Flip Vertical
    • Flip Horizontal
    • Zoom
    • Shrink
    • Crop
  2. Read and Display an Image File. Remove noisy pixels from the image using:
    • Horizontal and Vertical Neighbours
    • Diagonal Neighbours
    • All Neighbours

    Display the Output

  3. Read and display an image file. Apply the following operators on the image:
    • Identity Operator
    • Inverse Operator
    • Threshold Operator, assume threshold value p1
    • Inverted Threshold Operator, assume threshold value p1
    • Binary Threshold Interval Operator, assume threshold values p1 and p2
    • Inverted Binary Threshold Operator, assume threshold values p1 and p2
    • Grey Scale Threshold Operator, assume threshold values p1 and p2
    • Inverted Grey Scale Threshold Operator, assume threshold values p1 and p2. Find the r=effect of varying the values of p1 and p2
  4. Read and Display an Image File. Calculate the Histogram of the image, [Try to display the histogram on the screen]. Calculate the following Characteristic Properties of the image and display the calculated values on the screen: 
    • Brightness
    • Contrast
    • Entropy
    • Skewness
    • Spatial Frequency
    • Visibility

    Vary the values of brightness and contrast and study their effects on the image

  5. Read and display an image file. Apply Enhancement algorithm using:
    • Linear Contrast Stretching
    • Non-Linear Contrast Stretching- Exponential
    • Non-Linear Contrast Stretching- Logarithmic.
  6. An image matrix of 8-bit resolution is given below:

    Change the resolution of the image from 8 bit to 5 bit and 3 bit.
  7. Consider the 5x5 image matrix given below and perform the Zooming of the Image Matrix by using nearest neighbour and bilinear interpolation. Assume zooming factor as 3.
  8. Consider the 10 x10 matrix given below. Perform the Shrinking of the matrix the size of the output matrix should be half of the original matrix.

  9. Consider the image matrix given below. Perform the Linear, Logarithmic and Exponential stretching on it. Draw the input and output histogram and compare the results.

  10. Consider the following 4 x 4 (16 level) matrix given in the Figure below:


    Apply the Piecewise Algorithm on image segment with following parameters
    lmin 0 and lmax 15, a=2 and b=8. Compare the input and output matrices by drawing the histograms.
  11.  An 8-level image has grey level distribution shown in table below. rk indicates gray
    level of the image. Total no. of pixel = 36.

    • Plot the histogram
    • Calculate PDF and CDF of the histogram
    • Find the new histogram based on histogram equalization
  12.  An 8-level image has grey level distribution shown in table below. rk indicates gray level of the image. Total no. of pixel = 4096.

    • Plot the histogram
    • Calculate PDF and CDF of the histogram
    • Find the new histogram based on histogram equalization
  13.  An 8-level image has grey level distribution shown in table below. rk indicates gray level of the image. Total no. of pixel = 4096.

    • Plot the histogram
    • Calculate PDF and CDF of the histogram
    • Find the new histogram based on histogram equalization
  14.  An 8-level image has grey level distribution shown in table below indicates gray level of the image. Total no. of pixel = 225.

    • Plot the histogram
    • Calculate PDF and CDF of the histogram
    • Find the new histogram based on histogram equalization

Module IV:

  1. Finding out the minimum number of frames per second for smooth motion of objects in video sequence.
  2. Constructing video data considering minimum number of frames, use of interpolation methods for constructing smooth video sequences.
  3. Methods to remove noise in video sequence through averaging methods
  4. Use of suitable filtering methods for eliminating unwanted information in video sequences.
  5. Finding DC component of video sequences through Fourier methods and differentiating between magnitude and phase information of objects with motion and without motion.

Module V:

  1. Acquiring data form different sensors and digital cameras for processing.
  2. Conversion of video data into different formats.
  3. Study of impact of color conversion on information.
  4. Study of impact of information content on storage space with different video formats.
  5. Estimating storage requirements of video data for surveillance applications, medical applications.
  6. Study on importance of video indexing for video processing.

Project:

Mini project on image and video processing.

TEXT BOOK:

  1. Rafael C Gonzalez and Richard E Woods, “Digital Image Processing’’, Pearson Education, 3rd Edition, 2009.
  2. Chanda, D Dutta Majumder, “Digital Image Processing and Analysis’’, Prentice-Hall, India, 2002
  3. Video Processing and Communications" by Yao Wang, Joern Ostermann, and Ya-Qin Zhang, Prentice Hall
  4. Digital Video Processing" by M. Tekalp
  5. Handbook of Image and Video processing - Al Bovik (Alan C Bovik), Academic Press,

REFERENCE BOOKS:

  1. “Fundamentals of Digital Image Processing”, Anil K. Jain, PHI, 1995.
  2. “Digital Image Processing”, William. K.Pratt, Wiley Interscience, 2nd Ed, 1991.
  3. Richard Szeliski, Computer Vision: Algorithms and Applications, Springer-Verlag London Limited 2011

Course - 2: Advanced Video Processing and Its  Applications -- (3-3-0)

Course - 3: Machine Learning and ANN -- (2-2-0)

Course - 4: Embedded Hardware Board -- (2-2-0)

Course - 5: Control System -- (2-2-0)

Course - 6: Project -- (0-0-6)

Our Main Teachers

Dr. Jharna Majumdar joined as Prof. of Eminence, School of Engineering & Technolog[y [SoET], BBSR Campus on Januare 2021. Prior to this, Dr, Majumdar worked Dean R & D, Prof. Dept of M Tech Comp Sc & Engg. and Head, Centre for Robotics Research an d Centre For Space Research at the Nitte Meenakshi Institute […]