UCI/SMPTE Webinar 12 June 2012 by 2m18Zj

VIEWS: 0 PAGES: 7

									UCI/SMPTE Webinar 12 June 2012

Future of the Image and video compression
This webinar will address:

    1. Emerging Research and Standards in Next Generation Video Coding (HEVC)
    2. What can we do with this new standard?
    3. How to teach new standards and compression techniques?

A new High Efficiency Video Coding (HEVC) standard is on the way. HEVC provides a significantly higher
coding performance than MPEG-2 and AVC, especially for high-resolution video, and also includes
various technologies to enable parallel encoding/decoding and to simplify implementation. Although the
modules of HEVC exist in AVC, almost every module has been re-considered and many new and exciting
state-of-the-art coding tools have been introduced. Many ideas behind such tools, whether accepted or
not, are novel and inspiring to the research community. HEVC also presents many new challenges,
including encoding optimization, mode decision, rate-control, hardware design, and error concealment,
especially given its large and hierarchical block structure and significantly increased number of coding
parameters.


In the DTV, HEVC application in broadcasting depends of the infrastructure for the previous digital
broadcast standard. Are transmitter infrastructure and receiver base widely used? How soon will set-top
boxes be commercially evaluable, and at what price? Is it reasonable to use terrestrial broadcast
systems considering the level of their use and alternative DTV systems?

Dr. Isailovic developed through the years the special structured techniques for teaching MPEG
standards. This will be presented in this webinar.



Note: see (old) http://www.jritechnology.com/Instructor.pdf

                http://www.jritechnology.com/

Registration: http://unex.uci.edu/services/events/




                                                                                                      1
2
Quiz for UCI/SMPTE Webinar 12 June 2012
Part A: Mostly True/False
    1. Temporal sampling artifacts (e.g. backward-turning wagon wheels): once aliases have been
       created, they may not be removed from the reconstructed material, as they are within the band
       of frequencies of interest.
       True            False



2. Is JPEG video coding standard?

        YES     NO

3. Lossy compression in the JPEG is achieved through:

        a. quantization

        b. DCT transform

        c. Huffman coding

        d. Predictive coding



4. Standard TV screen aspect ratio is 4:3.

        True              False



5. Pixel aspect ratio is always the same as the screen/picture aspect ratio.

        True              False



6. Transform coding is compression.

        True              False

7. Quantization is a “many-to-one” process.

        True              False




                                                                                                   3
8. Motion estimation and motion compensation are both used in both - coder and decoder.

        True             False



9. In the 4:2:2 sampling structure color components are sub-sampled horizontally only.

        True             False

10. According the ITU-R (CCIR) 601 standard number of samples in the active line is the same for all
three major TV standards – NTSC, PAL, and SECAM.

        True             False



11. If the DCT coefficient is 166 and normalizing coefficient (quantizer step size) is 10, the quantized DCT
coefficient is 16 and dequantized DCT coefficient (reconstructed at the receiver) is 10x16 = 160.

        True             False



12. In essence DCT and entropy (Huffman) coding are lossless.

        True             False

13. In the JPEG, a channel error can not propagate through the whole image and have a
catastrophic effect.

        True             False

        .

14. For the digital video compression based codec all (analog TV) lines should be processed.

        True             False




15. The MPEG-1 encoder is not specified in the Standard.

        True             False



16. In the MPEG-2 sampling structure is always 4:2:0.

                                                                                                           4
        True             False



Part B: Problems
Problem 1: The CRT tube has a diameter d. For the given picture aspect ratio, PAR, find the area
utilization.

[Picture rate and Display rate]
Problem 2: Scene is recorded 24 pictures per second. In the scene (soccer) ball goes diagonally across the
screen in 5 pictures.

Show the ball position on the screen if the display rate is:

    A) 24 pictures per second
    B) 48 pictures per second, but each picture is displayed twice (to reduce flicker)




                                                                                                        5
Problem 3:


[Huffman decoding]

             Simbol          Code words   Probability
              b(1)      00                   0.23
              b(4)      010                  0.08
              b(0)      0110                 0.06
              b(5)      0111                 0.06
              b(2)      10                   0.30
              b(3)      110                  0.15
              b(6)      1110                 0.06
              b(7)      1111                 0.06



A) Decode the following stream:

0 1 0 0 0 1 1 0 0 0 1 0 1 1 1 1 0…

___ ___ ___ ___ ___ ___ ___ ___

                                                        6
B) Decode in the presence of noise, one bit is altered:

0 1 0 0 0 1           0    0 0 0 1 0 1 1 1 0…


___ ___ ___ ___ ___ ___ ___ ___

C) Decode in the presence of noise, one bit of your choice is altered:

0 1 0 0 0 1 1 0 0 0 1 0 1 1 1 0…


___ ___ ___ ___ ___ ___ ___ ___


Problem 4: How many fields per picture are there in progressive scanning?

Problem 5: A camera CCD sensor is a circle with radius R. The same size sensor is used to
generate pictures with the aspect ratio 4:3 and 16:9. Find the ratio of the two sensor areas – for
two aspect ratios (assume that each area is as large as possible for the given R).

Problem 6: If a video signal is coded 1 byte per sample, how many bytes per sample on average
are used for the following sampling structures:

       a) 4:4:4       b) 4:2:2       c) 4:1:1       d) 3:1:1       e) 4:2:0




                                                                                                 7

								
To top