Abstract
Frame rate up-conversion (FRUC) is a widely used technique to
enhance the visual quality of low frame rate video presented on liquid
crystal displays. One decisive factor for an effective FRUC algorithm
is to obtain true motion vectors (MVs) for motion compensated interpolation.
However, many current block-based motion estimation methods cannot
always provide reliable MVs. In this paper, a new FRUC algorithm employing
linear quadratic motion estimation and trilateral filtering motion
smoothing is presented to estimate true MVs. The MV obtained by minimizing
matching error is modeled as the measurement of true MV with noise.
To remove the noise, the measured MV is updated with the corresponding
predicted MV by using linear quadratic estimation. In motion smoothing
stage, a trilateral filter which considers local smoothness, neighboring
motion reliability as well as motion diversity at motion boundary,
is used to further improve the accuracy and density of motion vector
field (MVF). Experimental results demonstrate that the proposed algorithm
outperforms the conventional FRUC algorithms with better objective
and subjective image quality.
© 2015 IEEE
PDF Article
More Like This
Cited By
You do not have subscription access to this journal. Cited by links are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.
Contact your librarian or system administrator
or
Login to access Optica Member Subscription