{[ promptMessage ]}

Bookmark it

{[ promptMessage ]}

HW5 - total operation count depend on the block size 3...

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
Polytechnic University, Dept. Electrical and Computer Engineering EE4414 Multimedia Communication System II Fall 2005, Yao Wang ___________________________________________________________________________________ Homework 5 (Motion estimation for video coding) Reading Assignment: • Lecture slides • Excerpts from EL514-Multimedia Lab Manual on video coding Written Assignment (Quiz based on selected problems on 10/14) 1. Describe the major steps of EBMA algorithms with integer-pel accuracy, either in the form of a pseudo C-code, or matlab code, or flow graph. 2. For a video of 30 frame/second, 352x288/frame, what is the number of operations needed per second to accomplish EBMA if we use block size of 16x16, search range of –16 to 16? (count one subtraction and taking absolute value, and sum of two numbers as one operation). Does the
Background image of page 1
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: total operation count depend on the block size? 3. Describe the major steps of EBMA algorithms with half-pel accuracy, either in the form of a pseudo C-code, or matlab code, or flow graph. 4. What is the pros and cons of half-pel EBMA compared to integer-pel EBMA? 5. What will be the operation counts for half-pel accuracy EBMA for the same parameter setting of Problem 2? 6. What is motion-compensated temporal prediction? How is it used in video coding? How does it reduce the bit rate needed to code a video, compared to code each frame directly? 7. How would you apply EBMA to accomplish bi-directional motion compensated prediction?...
View Full Document

Report

{[ snackBarMessage ]}