Mesh based frame processing and applications
First Claim
1. A method of processing sequential frames of data, the method comprising repeating the following steps for successive frames of data:
- acquiring at least a reference frame containing data points and a current frame of data points;
identifying a set of anchor points in the reference frame;
assigning to each anchor point in the reference frame a respective motion vector that estimates the location of the anchor point in the current frame;
defining polygons formed of anchor points in the reference frame, each polygon containing data points in the reference frame, each polygon and each data point contained within the polygon having a predicted location in the current frame based on the motion vectors assigned to anchor points in the polygon;
for one or more polygons in the reference frame, adjusting the number of anchor points in the reference frame based on accuracy of the predicted locations of data points in the current frame; and
if the number of anchor points is increased by addition of new anchor points, then assigning motion vectors to the new anchor points that estimate the location of the anchor points in the current frame.
3 Assignments
0 Petitions
Accused Products
Abstract
A method of processing sequential frames of data comprises repeating the following steps for successive frames of data: acquiring at least a reference frame containing data points and a current frame of data points; identifying a set of anchor points in the reference frame; assigning to each anchor point in the reference frame a respective motion vector that estimates the location of the anchor point in the current frame; defining polygons formed of anchor points in the reference frame, each polygon containing data points in the reference frame, each polygon and each data point contained within the polygon having a predicted location in the current frame based on the motion vectors assigned to anchor points in the polygon; for one or more polygons in the reference frame, adjusting the number of anchor points in the reference frame based on accuracy of the predicted locations of data points in the current frame; and if the number of anchor points is increased by addition of new anchor points, then assigning motion vectors to the new anchor points that estimate the location of the anchor points in the current frame.
83 Citations
49 Claims
-
1. A method of processing sequential frames of data, the method comprising repeating the following steps for successive frames of data:
-
acquiring at least a reference frame containing data points and a current frame of data points;
identifying a set of anchor points in the reference frame;
assigning to each anchor point in the reference frame a respective motion vector that estimates the location of the anchor point in the current frame;
defining polygons formed of anchor points in the reference frame, each polygon containing data points in the reference frame, each polygon and each data point contained within the polygon having a predicted location in the current frame based on the motion vectors assigned to anchor points in the polygon;
for one or more polygons in the reference frame, adjusting the number of anchor points in the reference frame based on accuracy of the predicted locations of data points in the current frame; and
if the number of anchor points is increased by addition of new anchor points, then assigning motion vectors to the new anchor points that estimate the location of the anchor points in the current frame. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 37)
-
-
21. A method of processing sequential frames of data, the method comprising repeating the following steps for successive frames of data:
-
acquiring at least a reference frame containing data points and a current frame of data points;
identifying a set of anchor points in the reference frame;
assigning to each anchor point in the reference frame a respective motion vector that estimates the location of the anchor point in the current frame;
defining polygons in the reference frame, each polygon being defined by at least four anchor points;
for one or more polygons in the reference frame, adjusting the number of anchor points in the reference frame; and
if the number of anchor points is increased by addition of new anchor points, then assigning motion vectors to the new anchor points that estimate the location of the anchor points in the current frame. - View Dependent Claims (22, 23, 24, 25, 26, 27, 28, 29, 30, 38)
-
-
31. A method of processing sequential frames of data, the method comprising repeating the following steps for successive frames of data:
-
acquiring at least a reference frame containing data points and a current frame of data points;
identifying a set of anchor points in the reference frame, where the distribution of anchor points is based on texture of the reference frame;
assigning to each anchor point in the reference frame a respective motion vector that estimates the location of the anchor point in the current frame;
defining polygons in the reference frame, each polygon being defined by anchor points;
assigning tracking points to at least one object in the reference frame, each object being formed by one or more polygons; and
analyzing the object using the tracking points. - View Dependent Claims (32, 33, 34, 35, 36, 39)
-
-
40. A method of processing a frame of data, the method comprising the steps of:
-
dividing the frame of data into cells; and
processing the cells according to an order that preserves the proximity of the processed cells, such that neighboring cells are processed in close sequence to each other. - View Dependent Claims (41, 42, 43)
-
-
44. A method of tracking an object across frames of data, the method comprising the steps of:
-
identifying a set of n tracking points on the object; and
through successive frames, following different combinations of the set of n tracking points to track the object while specific ones of the n tracking points are not usable for tracking.
-
-
45. A method of analyzing frames of data from a source of frames, the method comprising the steps of:
-
identifying objects in the frames of data, where the objects comprise patches of data having similar motion; and
identifying an unusual condition as occurring when the object size corresponds to the frame size. - View Dependent Claims (46, 47)
-
-
48. A method of encoding a bit stream carrying a frame of data, the method comprising the steps of:
-
encoding a first portion of the frame of data using a first compression algorithm to generate a first code stream;
encoding a second portion of the frame of data using anchor points of a mesh, the anchor points having associated motion vectors to generate a second code stream; and
transmitting a bit stream containing the first code stream and the second code stream. - View Dependent Claims (49)
-
Specification