Docstoc

Enhanced Hexagonal Search for Fast Block Motion Estimation

Document Sample
Enhanced Hexagonal Search for Fast Block Motion Estimation Powered By Docstoc
					 Enhanced Hexagonal
 Search for Fast Block
  Motion Estimation

Authors:Ce Zhu, Xiao Lin,
        Lappui Chau, and Lai-Man Po
IEEE TRANSACTIONS ON CIRCUITS AND
SYSTEMS FOR VIDEO TECHNOLOGY,
OCTOBER 2004
              Outline
   Introduction
   HEXBS ( Hexagon-Based Search )
   Predictive HEXBS
   Fast Hexagonal Inner Search
   6-Side-Based Fast Inner Search
   Enhanced HEXBS Algorithm
   Experimental Results And Analysis
   Conclusion
              Introduction
   Fast block motion estimation process :
    STEP 1:low-resolution coarse search
            → To identify a small area where the
              best motion vector is expected to lie
    STEP 2:fine-resolution inner search
            → To select the best motion vector in
              the located small region
             Introduction
 Most motion estimation algorithms attempt to
  speed up the coarse search without considering
  accelerating the inner search
 Enhanced hexagonal search algorithm is
  proposed to improve the performance:
 (1) Reducing number of search points
 (2) Decrease the distortion
              Introduction
   The two-dimensional logarithmic search
   Three-step search (TSS)
   New three-step search (NTSS)
   Four-step search (4SS)
   Block-based gradient descent search (BBGDS)
   Simple and efficient search (SES)
   Diamond search (DS)
   The hexagonal search
   Hexagon-based search (HEXBS)
   Enhanced hexagon-based search
HEXBS
               HEXBS




Inner points in the hexagonal search pattern
The inner search for the HEXBS
 Using the shrunk hexagonal pattern covering the
  points 2, 4, 5, and 7
 A gradient scheme:
  Ex:Points 1 and 3 if point 2 wins in the last step
      of the HEXBS algorithm
                The flowchart of HEXBS




NHEXBS (mx , my) = 7 + 3n + 4
where n is the number of times of
low-resolution coarse search
           Predictive HEXBS
   The error distortion function has monotonic
    characteristic in a localized search area
   The motion vector of the current block is
    highly correlated to those of its neighboring
    blocks.
   The motion information of neighboring blocks
    can be utilized for prediction of a good starting
    point
           Predictive HEXBS
   Consider the upper and the left neighboring
    blocks
   Finding a good starting point using the
    neighboring motion vectors
   Normally finds better motion vectors than the
    original HEXBS scheme
Fast Hexagonal Inner Search

   Not a full inner search , only check a portion of
    the inner search points
   Strong correlation exists between the inner
    search points
   Based on the monotonic distortion characteristic
    in the localized area around the global minimum
6-Side-Based Fast Inner Search
   Group the search points in the six sides of the
    hexagon
   Define a group distortion by summing the
    distortions of all the points within the group
   We focus the inner search just in the region
    near to the group with the smallest group
    distortion
   For different groups (sides) in different
    locations, we have different number of inner
    search points
Enhanced HEXBS Algorithm
 HEXBS incorporate the 6-side-based fast
  inner search scheme
 Moreover , incorporate the Predictive HEXBS

 The reduction of number of search points for
  the enhanced HEXBS algorithm :
  (1) The prediction for a good starting point
      using the predictive HEXBS,
  (2) The fast inner search.
Experimental Results And Analysis
                                     TABLE III




           N 2 - N1
%SIR  (            ) 100% , where Ni is the number of search points used in the Method i
             N1

             MSE 2 - MSE 1
%D MSE  (                 ) 100%    ,where MSEi is the distortions for Methods i
               MSE 1
                 Conclusion
   Enhanced HEXBS speeds up the motion
    estimation and decreases distortions
   Only part of the inner points will be evaluated
   Enhanced HEXBS algorithm outperforms the
    original HEXBS

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:10
posted:2/17/2012
language:
pages:20