External vectors in a motion estimation system
First Claim
1. A machine-implemented method for motion estimation in a sequence of video images, comprising:
- subdividing each field or frame of a sequence of video images into a plurality of blocks;
assigning to each block in each video field or frame a set of candidate motion vectors;
for each block in a current video field or frame, selecting a vector that produces a best match to a block in a previous video field or frame, from the set of candidate motion vectors assigned to that block; and
forming a motion vector field for the current video field or frame using the selected vector for each block,wherein the set of candidate motion vectors assigned to a block include one or more candidate vectors derived from an external source vector field,and wherein each candidate motion vector from an external source is assigned a weighting value, that biases towards or away from the selection of that candidate motion vector.
1 Assignment
0 Petitions
Accused Products
Abstract
A method and apparatus for motion estimation in a sequence of video images is provided. Each field or frame in a sequence of video images is sub-divided into a plurality of blocks. Each block in each video field or frame has assigned to it a set of candidate motion vectors. The vector which produces a best match to a block in a previous field or frame, from the set of candidate motion vectors, is assigned to that block thus forming the motion vector field for the current video field or frame using the selected vector. The set of candidate motion vectors assigned to a block include one or more candidate vectors derived from an external source vector field and each such vector from an external source vector field is assigned a weighting that biases towards or away from the selection of that vector.
24 Citations
36 Claims
-
1. A machine-implemented method for motion estimation in a sequence of video images, comprising:
-
subdividing each field or frame of a sequence of video images into a plurality of blocks; assigning to each block in each video field or frame a set of candidate motion vectors; for each block in a current video field or frame, selecting a vector that produces a best match to a block in a previous video field or frame, from the set of candidate motion vectors assigned to that block; and forming a motion vector field for the current video field or frame using the selected vector for each block, wherein the set of candidate motion vectors assigned to a block include one or more candidate vectors derived from an external source vector field, and wherein each candidate motion vector from an external source is assigned a weighting value, that biases towards or away from the selection of that candidate motion vector. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20)
-
-
21. Apparatus for motion estimation in a sequence of video images comprising:
-
a source of pluralities of blocks, each plurality from a respective subdivided field or frame of a sequence of video images; a generator of internal motion vector candidates; one or more external sources of candidate motion vectors; a memory coupled to the generator of internal motion vector candidates and to the one or more external sources of candidate motion vectors, and storing a motion vector candidate list, comprising internal motion vector candidates and one or more motion vectors from the one or more external sources of candidate motion vectors; a selector configured, for each block in a current video field or frame, to select a vector that produces a best match to a block in a previous video field or frame of the sequence of video images, from a set of candidate motion vectors assigned to that block from the motion vector candidate list; and one or more of a deinterlacer and a frame rate converter configured to form a motion vector field for the current video field or frame using the selected vector for each block, wherein each candidate motion vector from the external source is assigned a weighting value that biases towards or away from the selection of that candidate motion vector. - View Dependent Claims (22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36)
-
Specification