Claims Analysis
Total Claims: 22
Claim 1
Independent Claim
89 words
A method of estimating motion of at least one picture portion in an image of a picture signal, comprising the steps of: determining a number of candidate motion vectors starting from a corresponding number of starting vectors, and selecting a motion vector out of the determined candidate motion vectors, characterized by taking as components of the starting vectors, components of candidate motion vectors already previously determined, which candidate motion vectors correspond respectively to the starting vectors, and which candidate motion vector components correspond respectively to the starting vector components.
Claim 2
Independent Claim
89 words
A method of estimating motion of at least one picture portion in an image of a picture signal, comprising the steps of: determining a number of candidate motion vectors starting from a corresponding number of starting vectors, and selecting a motion vector out of the determined candidate motion vectors, characterized by taking as components of the starting vectors, components of candidate motion vectors already previously determined, which candidate motion vectors correspond respectively to the starting vectors, and which candidate motion vector components correspond respectively to the starting vector components.
Claim 3
Dependent Claim
63 words
A method as claimed in claim 1, wherein picture elements of each of a plurality of picture portions are divided into a first and a second group, respectively, by sub-sampling the picture portion in a first and a second phase, a first candidate motion vector being determined for the first group and a second candidate motion vector being determined for the second group.
Claim 4
Dependent Claim
153 words
A method as claimed in claim 1, wherein a first component of a first starting vector is taken to be equal to a first component of a first candidate motion vector already previously determined for a first picture portion, a second component of said first starting vector is taken to be equal to a second component of a second candidate motion vector already previously determined for a second picture portion, while the first and second picture portions do not coincide, and in that a first component of a second starting vector is taken to be equal to a first component of a third candidate motion vector already determined for a third picture portion, a second component of said second starting vector is taken to be equal to a second component of a fourth candidate motion vector already determined for a fourth picture portion, while the third and fourth picture portions do not coincide.
Claim 5
Dependent Claim
24 words
A method as claimed in claim 3, wherein the third and the fourth picture portions coincide with the second and first picture portions, respectively.
Claim 6
Dependent Claim
57 words
A method as claimed in claim 3 wherein the first and the second picture portions have a smaller y-coordinate than the picture portion for which the motion is being estimated, the first picture portion having a smaller x-coordinate and the second picture portion a larger x-coordinate than the picture portion for which the motion is being estimated.
Claim 7
Dependent Claim
28 words
A method as claimed in claim 2, wherein the picture portions in a picture for each row of picture portions are picture element blocks which are shiftedly stacked.
Claim 8
Dependent Claim
36 words
A method as claimed in claim 1, wherein the candidate motion vectors are determined in accordance with a criterion that provides that only a candidate motion vector which differs from the starting vector can be determined.
Claim 9
Dependent Claim
17 words
A method as claimed in claim 1, wherein components of the selected vectors are spatially low-pass filtered.
Claim 10
Dependent Claim
84 words
A method as claimed in claim 8, wherein said low-pass filtering step includes the steps of recursively filtering said selected vector components to obtain output components; comparing said output components to said selected vector components to obtain a control signal indicating whether said selected vector components differ more than a given threshold from said output vector components; and disabling said recursively filtering step if said control signal indicates that said selected vector components differ more than said given threshold from said output vector components.
Claim 11
Dependent Claim
16 words
A method as claimed in claim 8, wherein the low-pass filtered vector components are spatially sub-sampled.
Claim 12
Dependent Claim
56 words
A method as claimed in claim 1, wherein, previous to the motion estimation the picture signal is temporally low-pass filtered utilizing a temporal filter having coefficients (1/4, 1/2, 1/4), wherein from field to field, lines from both adjacent fields, located alternately above or below a line of a present field are multiplied by the coefficient 1/4.
Claim 13
Dependent Claim
56 words
A method as claimed in claim 1, wherein said candidate motion vector determining step includes the steps of: taking the median of said picture signal, a field-delayed picture signal and a field- and line-delayed picture signal to obtain a filtered signal; and estimating said candidate motion vectors based on said picture signal and said filtered signal.
Claim 14
Dependent Claim
28 words
A method as claimed in claim 3, wherein the picture portions in a picture for each row of picture portions are picture element blocks which are shiftedly stacked.
Claim 15
Dependent Claim
28 words
A method as claimed in claim 4, wherein the picture portions in a picture for each row of picture portions are picture element blocks which are shiftedly stacked.
Claim 16
Dependent Claim
28 words
A method as claimed in claim 5, wherein the picture portions in a picture for each row of picture portions are picture element blocks which are shiftedly stacked.
Claim 17
Dependent Claim
36 words
A method as claimed in claim 15, wherein the candidate motion vectors are determined in accordance with a criterion that provides that only a candidate motion vector which differs from the starting vector can be determined.
Claim 18
Dependent Claim
17 words
A method as claimed in claim 16, wherein components of the selected vectors are spatially low-pass filtered.
Claim 19
Dependent Claim
57 words
A method as claimed in claim 10, wherein previous to the motion estimation the picture signal is temporally low-pass filtered utilizing a temporal filter having the coefficients 1/4, 1/2, 1/4, wherein from field to field, lines from both adjacent fields, located alternately above or below a line of a present field are multiplied by the coefficient 1/4.
Claim 20
Independent Claim
104 words
A method of estimating motion per picture portion in an image of a picture signal wherein, starting from a starting vector, a motion vector is determined according to a previously defined criterion, characterized in that a first component of the starting vector is taken to be equal to a first component of a first already previously estimated motion vector, belonging to a first picture portion, and a second component of the starting vector is taken to be equal to a second component of a second, already previously estimated motion vector, belonging to a second picture portion, said first and second picture portions not coinciding.
Claim 21
Independent Claim
104 words
A method of estimating motion per picture portion in an image of a picture signal wherein, starting from a starting vector, a motion vector is determined according to a previously defined criterion, characterized in that a first component of the starting vector is taken to be equal to a first component of a first already previously estimated motion vector, belonging to a first picture portion, and a second component of the starting vector is taken to be equal to a second component of a second, already previously estimated motion vector, belonging to a second picture portion, said first and second picture portions not coinciding.
Claim 22
Dependent Claim
99 words
A method as claimed in claim 19, wherein for each picture portion, in addition to said estimation, a second motion estimation is also performed, wherein for the second motion estimation components of a second starting vector are assumed to be equal to respective corresponding components of third and fourth, respectively, already previously estimated motion vectors associated with third and fourth, non-coinciding picture portions, respectively, and starting from the second starting vector a second motion vector is determined in accordance with the said criterion, whereafter a best motion vector is selected from said first and said second determined motion vectors.