Digital Video Processing

Teacher

N.V.S. Shankar

Category

Course Attendees

Still no participant

Course Reviews

Still no reviews

Course Name: Digital Video Processing
Course Code: CUML2012 (0-2-1)

Course Objectives:

  • Introduce students to the basics of digital video acquisition, including the principles of color video, video camera, and video display.
  • Provide students with an understanding of digital video compression metrics, digital video storage precisions, and the significance of video compression.
  • Make students aware of video editing techniques that can be used for ML applications

Course Outcomes

  • Analyze the principles of color video, video camera, and video display to identify the best practices for digital video acquisition.
  • Evaluate the significance of video compression and its impact on digital video storage precisions and data rate.
  • Create a video editing pipeline that incorporates machine learning techniques to enhance video quality and performance.

UNIT 1:
Fundamentals of Video Processing: Digital Video Acquisition, Principles of Color Video, Video Camera, Video Display, Analog Vs Digital Video: Progressive Vs Interlaced scans, Signal, Bandwidth Characterization of a Digital Video Signal.
Practice:

  • Read and play video files
  • Extract frames from video files
  • Combine frames to create a video file

UNIT 2:

Fourier Analysis of Digital Video Signals: Spatial and Temporal resolution, Fourier Analysis of Digital Video Signals, Spatial-Temporal Sampling: Temporal Frequency Response and Flicker Perception. Spatial Frequency Response, Spatiotemporal Frequency Response, Smooth Pursuit Eye Movement

Practice:

  • Applying fourier transformation on video
  • Time domain analysis
  • Frequency domain Analysis

UNIT 3:
Digital Video Formats: Significance of Video Formatting, Data rate and bandwidth trade-off
File Formats: MP4, MOV, WMV, AVCHD, FLV, AVI, WebM, MKV
Digital Video Compression Standards: Digital Video Compression Metrics, Digital Video Storage Precisions, Significance of Video compression, Video Compression Codec’s: Motion JPEG, JPEG 2000, H.264/MPEG-4 AVC, VP8, HEVC, H.265 High Efficiency Video Codec.

Practice:

  • Conversion of video files from one format to another using Motion JPEG, MPEG-4, H.265 Codec

UNIT 4:
Digital Video Editing Basics: Video Editing Types- Online, Offline, Linear, Non-linear, Assemble, Insert, Rough-cut, Video Shot Transition Effects: Cut, Fade, Wipe, Dissolve, B-roll, Video Shot Boundary Detection Methods: pixel differences, statistical differences, histogram comparisons, edge differences and motion vectors. Video Shot Detection Performance Metrics: ROC Curves, Recall, Precision, F-Measure

Practice:

  • Video Shot Detection using pixel Difference
  • Video Shot Detection using Histogram based methods
  • Video Shot Detection using Edge based methods
  • Video Shot Detection using Motion Vectors

Text Books:

  1. Rafael C Gonzalez and Richard E Woods, “Digital Image Processing’’, Pearson Education, 3rd Edition,
  2. Handbook of Image and Video processing - Al Bovik (Alan C Bovik), Academic Press,

Reference Books:

  1. Fundamentals of Digital Image Processing”, Anil Jain, PHI, 1995.
  2. Digital Image Processing, William. K.Pratt, Wiley Interscience, 2nd Ed, 1991.

Complete Lab Manual: DVP Lab Manual

Session Plan (3hr practice session)
Session 1: Fundamentals of Video Processing, Analog Vs Digital Video
Session 2: Read and play video files, Extracting Frames
Session 3: Combine frames to create a video file
Session 4: Spatial-Temporal Sampling, Time domain analysis
Session 5: Fourier Analysis of Digital Video Signals - Spatiotemporal Analysis
Session 6: Frame Interpolation to Reduce Flicker
Session 7: Motion Analysis of Video
Session 8: Smooth Pursuit Eye Movement
Session 9: Digital Video Formats and Digital Video Compression Standards, Conversion of video files from one format to another
Session 10: Using Motion JPEG, MPEG-4 and H.265 Codecs
Session 11: Video Shot Detection using pixel difference and Histogram Difference
Session 12: Video Shot Detection using Edge based methods and Motion Vectors

Our Main Teachers

N.V.S. Shankar

Associate Professor
VIEW PROFILE