Method for assessing the quality of a distorted version of a frame sequence
View Patent ↗A method for assessing the quality of a distorted version of a frame sequence includes the steps of determining a last spatial distortion by comparing a block of a last frame (I 6 ) of the sequence with a corresponding block of the distorted version of the last frame, determining, in a last-but-one frame, a best-matching block matching said block of the last frame best, determining a last-but-one spatial distortion by comparing the determined best-matching block of the last-but-one frame with a corresponding block of the distorted version of the last-but-one frame, determining a spatio-temporal perceptual distortion value associated with said block using said determined distortions and using the determined spatio-temporal perceptual distortion value for assessing the quality.
1. A method for assessing quality of a distorted version of a frame sequence via a hardware device, said method comprising:
for each block to be assessed of a last frame of said frame sequence,
a) determining a last spatial distortion by comparing said block to be assessed with a corresponding block of said distorted version of said last frame of said distorted version of said frame sequence, said corresponding block of said distorted version of said last frame being positioned in said distorted version of said last frame at a same position as said block to be assessed;
b) determining a last-but-one spatial distortion by comparing a best-matching block of a last-but-one frame with a corresponding block of said distorted version of said last-but-one frame, said best-matching block of said last-but-one frame being determined using motion estimation; and
c) determining a spatio-temporal perceptual distortion value associated with said block to be assessed using said last spatial distortion and said last-but-one spatial distortion for assessing said quality of said block to be assessed;
d) determining, in a last-but-two frame, a further best-matching block matching said best matching block of said last-but-one frame best;
e) determining a last-but-two spatial distortion by comparing said determined further best-matching block of said last-but-two frame with a corresponding block of said distorted version of said last-but-two frame; and
iterating steps d) and e) through a number of frames preceding said last frame; and
wherein said spatio-temporal perceptual distortion value is determined using a weighted average of all determined spatial distortions and distortion gradients between consecutive frames are used for determining a weight for weighting said spatio-temporal perceptual distortion value.
2. The method according to claim 1 , further comprising:
determining whether a first distortion gradient meets or exceeds a threshold; and
if said first distortion gradient meets or exceeds said threshold, continuing iterating steps d) and e) through a further number of frames.
3. The method according to claim 1 , further comprising:
counting a number of sign changes between determined consecutive distortion gradients; and
convolving said number of sign changes, wherein said weight is further proportional to said convolved number of sign changes.