Search icone
Search and publish your papers
Our Guarantee
We guarantee quality.
Find out more!

A new motion compensation technique for efficient video coding

Or download with : a doc exchange

About the author


About the document

Published date
documents in English
term papers
6 pages
0 times
Validated by
0 Comment
Rate this document
  1. Abstract
  2. Introduction
  4. Proposed motion compensation model
    1. Algorithm motion compensation
  5. Analysis and experimental results
  6. Conclusion
  7. References

Motion estimation and motion compensation are two interrelated and obviously important tasks in the total video coding process. Out of several motion estimation methods block matching algorithm (BMA) is the simplest, hence rightly adopted by several MPEGs and H.26X standards. Motion compensation is always associated with the motion estimation technique, which evaluate the error or residual image (block-wise) between the original frame and predicted frame. But the associated overhead of the error image requires more storage/channel bandwidth. In this paper we present an efficient motion compensation technique based on a predefined threshold value, we termed it as d-threshold value (?). For applying our motion compensation technique we choose a recently reported modification of Hexagonal BMA, Threshold based Hexagonal (TBHEX) BMA [1] which has proved its efficiency in reducing the number of search points while maintains a good reconstructed image quality. Experimental results show that our approach requires maximum 29.60% of total block's motion compensation on standard video bench-marks.

[...] For applying our motion compensation approach we choose a recently reported Threshold-based Hexagonal BMA modification of Hexagonal BMA), which has less number of search points than that of Hexagonal BMA while maintains the quality of the Hexagonal search. Rest of the paper is organized as follows. In Section II we discuss the TBHEX BMA. Our proposed motion compensation model is presented in Section III. Experimental results and analysis are presented in Section IV. Finally the paper is concluded in Section V. [...]

[...] Such situation may arise due to the complicated motion like zooming, occlusion, rotation and arrival of new object Frame replenishment is a good technique which checks the active regions (which do not have significant correlation with the reference frame) of an image and stores/transmits the information of the all active regions while the passive regions PSNR d 2 ) = 10 log10 (which have significant correlation with reference frame) are represented by motion vectors. But the frame replenishment technique requires massive data storage or large bandwidth channel for storing or transmitting videos. [...]

[...] For high motion video like Caltrain, Football, Tennis %EBI are and respectively. Table 1 shows the results on the %EBI on all the 9 video bench-marks. Table 2 denotes the %DMSE (Decrement MSE) defined by the Equation 7. MAX MSE CURRENTMSE % DMSE = 100% MAX MSE Where MAXMSE denotes the maximum value of the MSE among all the 9 d-threshold values and CURRENTMSE denotes the MSE value for any d-threshold value among the 9 dthreshold values. Table 2 indicates that for high motion video %DMSE reduces (football) (tennis) EBI = vg. [...]

Recent documents in computer science category

Reconstructing householder vectors from tall-skinny QR

 Science & technology   |  Computer science   |  Presentation   |  04/21/2017   |   .doc   |   4 pages

Software requirement development - The airline ticketing reservations software systems

 Science & technology   |  Computer science   |  Presentation   |  01/30/2017   |   .doc   |   3 pages