Apparatus and method for capturing the expression of a performer
First Claim
Patent Images
1. A method, comprising:
- specifying a predefined facial curve pattern comprising a mesh of intersecting curves, wherein each individual curve in the mesh is uniquely identifiable;
applying the predefined facial curve pattern to a performer'"'"'s face;
storing the predefined facial curve pattern as curve data on a storage device, the curve data usable by a camera controller;
processing sequences of real-time visual images taken of said curves as said curves move during a motion capture session, including, processing visual images taken from regions of the curves residing between intersection points amongst the curves, said processing including correlating each of the captured real-time video images with the curve data read by one or more of the camera controllers to uniquely identify each of the curves, and intersection points between the curves, the operation of processing further comprising;
(i) identifying a first set of positions of the curves, the first set of positions of the curves forming a first set of surface tiles having curved edges; and
(ii) identifying a second set of positions of the curves, the second set of positions of the curves forming a second set of surface tiles having curved edges;
generating motion data from the recorded first and second sets of positions of the curves captured in the sequences of real-time visual images, the motion data representing movement of the performer'"'"'s face during the motion capture session; and
storing the motion data on a storage device, the motion data usable to graphically reconstruct movement of the curves during the motion capture session;
wherein a first set of one or more cameras capture a first sequence of real-time video images of a adjusted, predefined facial curve pattern from a first orientation as the performer makes facial movements during a performance and a second set of one or more cameras capture a second sequence of real-time video images of the adjusted, predefined facial curve pattern from a second orientation as the performer makes facial movements during a performance; and
wherein specifying the predefined facial curve pattern comprises defining a plurality of curve types and categorizing each of the uniquely identifiable curves as one of the defined curve types, the curve types including contours, radials, and transition curves;
wherein the predefined facial curve pattern is generated such that no curve can intersect another curve of the same curve type.
20 Assignments
0 Petitions
Accused Products
Abstract
A method is described comprising: applying a series of curves on specified regions of a performer'"'"'s face; tracking the movement of the series of curves during a motion capture session; and generating motion data representing the movement of the performer'"'"'s face using the tracked movement of the series of curves.
-
Citations
21 Claims
-
1. A method, comprising:
-
specifying a predefined facial curve pattern comprising a mesh of intersecting curves, wherein each individual curve in the mesh is uniquely identifiable; applying the predefined facial curve pattern to a performer'"'"'s face; storing the predefined facial curve pattern as curve data on a storage device, the curve data usable by a camera controller; processing sequences of real-time visual images taken of said curves as said curves move during a motion capture session, including, processing visual images taken from regions of the curves residing between intersection points amongst the curves, said processing including correlating each of the captured real-time video images with the curve data read by one or more of the camera controllers to uniquely identify each of the curves, and intersection points between the curves, the operation of processing further comprising; (i) identifying a first set of positions of the curves, the first set of positions of the curves forming a first set of surface tiles having curved edges; and (ii) identifying a second set of positions of the curves, the second set of positions of the curves forming a second set of surface tiles having curved edges; generating motion data from the recorded first and second sets of positions of the curves captured in the sequences of real-time visual images, the motion data representing movement of the performer'"'"'s face during the motion capture session; and storing the motion data on a storage device, the motion data usable to graphically reconstruct movement of the curves during the motion capture session; wherein a first set of one or more cameras capture a first sequence of real-time video images of a adjusted, predefined facial curve pattern from a first orientation as the performer makes facial movements during a performance and a second set of one or more cameras capture a second sequence of real-time video images of the adjusted, predefined facial curve pattern from a second orientation as the performer makes facial movements during a performance; and wherein specifying the predefined facial curve pattern comprises defining a plurality of curve types and categorizing each of the uniquely identifiable curves as one of the defined curve types, the curve types including contours, radials, and transition curves; wherein the predefined facial curve pattern is generated such that no curve can intersect another curve of the same curve type. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10)
-
-
11. A method for performing motion capture using curves comprising:
-
specifying a predefined facial curve pattern comprising a set of intersecting curves, wherein each individual curve in the set is uniquely identifiable; adjusting the predefined facial curve pattern based on a topology of a performer'"'"'s face and applying the adjusted, predefined facial curve pattern to the performer'"'"'s face; storing the adjusted, predefined facial curve pattern on a storage device, the curve data usable by a camera controller; tracking the movement of the series of curves of the adjusted, predefined facial curve pattern on the performer'"'"'s face during a motion capture session by performing the operations of; capturing sequences of real-time video images of the adjusted, predefined facial curve pattern as the performer makes facial movements during a performance; and correlating each of the captured real-time video images with the curve data read by one or more of the camera controllers to uniquely identify each of the curves and intersection points between the curves; generating motion data representing movement of each of the curves and the intersection points between the curves across the sequences of real-time video images; and storing the motion data on a storage device, the motion data usable for reconstructing the facial movements of the performer during the performance; wherein a first set of one or more cameras capture a first sequence of real-time video images of the adjusted, predefined facial curve pattern from a first orientation as the performer makes facial movements during a performance and a second set of one or more cameras capture a second sequence of real-time video images of the adjusted, predefined facial curve pattern from a second orientation as the performer makes facial movements during a performance; wherein generating the predefined facial curve pattern comprises defining a plurality of curve types and categorizing each of the uniquely identifiable curves as one of the defined curve types, the curve types including contours, radials, and transition curves; wherein the predefined curve pattern is generated such that no curve can intersect another curve of the same curve type. - View Dependent Claims (12, 13, 14, 15, 16)
-
-
17. A method, comprising:
-
specifying a predefined facial curve pattern comprising a mesh of intersecting curves, wherein each individual curve in the mesh is uniquely identifiable; applying the predefined facial curve pattern to a performer'"'"'s face; storing the predefined facial curve pattern as curve data on a storage device, the curve data usable by a camera controller; processing sequences of real-time visual images taken of said curves as said curves move during a motion capture session, including, processing visual images taken from regions of the curves residing between intersection points amongst the curves, said processing including correlating each of the captured real-time video images with the curve data read by one or more of the camera controllers to uniquely identify each of the curves, and intersection points between the curves, the operation of processing further comprising; (i) identifying a first set of positions of the curves, the first set of positions of the curves forming a first set of surface tiles having curved edges; and (ii) identifying a second set of positions of the curves, the second set of positions of the curves forming a second set of surface tiles having curved edges; generating motion data from the recorded first and second sets of positions of the curves captured in the sequences of real-time visual images, the motion data representing movement of the performer'"'"'s face during the motion capture session; and storing the motion data on a storage device, the motion data usable to graphically reconstruct movement of the curves during the motion capture session; wherein a first set of one or more cameras capture a first sequence of real-time video images of a adjusted, predefined facial curve pattern from a first orientation as the performer makes facial movements during a performance and a second set of one or more cameras capture a second sequence of real-time video images of the adjusted, predefined facial curve pattern from a second orientation as the performer makes facial movements during a performance; and wherein specifying the predefined facial curve pattern comprises defining a plurality of curve types and categorizing each of the uniquely identifiable curves as one of the defined curve types, the curve types including contours, radials, and transition curves; wherein the curve data comprises;
a list of curves organized by type, with each curve having a unique identifier associated therewith;
for each curve, an ordered list of other curves that it intersects along its length; and
a list of surface patches, each defined by the curves that make up its sides.
-
-
18. A method, comprising:
-
specifying a predefined facial curve pattern comprising a mesh of intersecting curves, wherein each individual curve in the mesh is uniquely identifiable; applying the predefined facial curve pattern to a performer'"'"'s face; storing the predefined facial curve pattern as curve data on a storage device, the curve data usable by a camera controller; processing sequences of real-time visual images taken of said curves as said curves move during a motion capture session, including, processing visual images taken from regions of the curves residing between intersection points amongst the curves, said processing including correlating each of the captured real-time video images with the curve data read by one or more of the camera controllers to uniquely identify each of the curves, and intersection points between the curves, the operation of processing further comprising; (i) identifying a first set of positions of the curves, the first set of positions of the curves forming a first set of surface tiles having curved edges; and (ii) identifying a second set of positions of the curves, the second set of positions of the curves forming a second set of surface tiles having curved edges; generating motion data from the recorded first and second sets of positions of the curves captured in the sequences of real-time visual images, the motion data representing movement of the performer'"'"'s face during the motion capture session; and storing the motion data on a storage device, the motion data usable to graphically reconstruct movement of the curves during the motion capture session; wherein a first set of one or more cameras capture a first sequence of real-time video images of a adjusted, predefined facial curve pattern from a first orientation as the performer makes facial movements during a performance and a second set of one or more cameras capture a second sequence of real-time video images of the adjusted, predefined facial curve pattern from a second orientation as the performer makes facial movements during a performance; and wherein specifying the predefined facial curve pattern comprises defining a plurality of curve types and categorizing each of the uniquely identifiable curves as one of the defined curve types, the curve types including contours, radials, and transition curves; wherein generating motion data further comprises;
quantizing each of the curves into a plurality of line segments between each of the intersection points; and
defining a plurality of triangles using the endpoints of each of the line segments and the intersection points of the curves as vertices of the triangles.
-
-
19. A method for performing motion capture using curves comprising:
-
specifying a predefined facial curve pattern comprising a set of intersecting curves, wherein each individual curve in the set is uniquely identifiable; adjusting the predefined facial curve pattern based on a topology of a performer'"'"'s face and applying the adjusted, predefined facial curve pattern to the performer'"'"'s face; storing the adjusted, predefined facial curve pattern on a storage device, the curve data usable by a camera controller; tracking the movement of the series of curves of the adjusted, predefined facial curve pattern on the performer'"'"'s face during a motion capture session by performing the operations of; capturing sequences of real-time video images of the adjusted, predefined facial curve pattern as the performer makes facial movements during a performance; and correlating each of the captured real-time video images with the curve data read by one or more of the camera controllers to uniquely identify each of the curves and intersection points between the curves; generating motion data representing movement of each of the curves and the intersection points between the curves across the sequences of real-time video images; and storing the motion data on a storage device, the motion data usable for reconstructing the facial movements of the performer during the performance; wherein a first set of one or more cameras capture a first sequence of real-time video images of the adjusted, predefined facial curve pattern from a first orientation as the performer makes facial movements during a performance and a second set of one or more cameras capture a second sequence of real-time video images of the adjusted, predefined facial curve pattern from a second orientation as the performer makes facial movements during a performance; wherein generating the predefined facial curve pattern comprises defining a plurality of curve types and categorizing each of the uniquely identifiable curves as one of the defined curve types, the curve types including contours, radials, and transition curves; wherein the predefined facial curve pattern defines different colors for one or more of the curves, the different colors usable to distinguish and uniquely identify each of the curves, wherein the curve data indicates a particular color for each of the uniquely identifiable curves, and wherein correlating comprises matching the color of each of the curves captured in the sequences of real-time video images with the color indicated in the curve data.
-
-
20. A method for performing motion capture using curves comprising:
-
specifying a predefined facial curve pattern comprising a set of intersecting curves, wherein each individual curve in the set is uniquely identifiable; adjusting the predefined facial curve pattern based on a topology of a performer'"'"'s face and applying the adjusted, predefined facial curve pattern to the performer'"'"'s face; storing the adjusted, predefined facial curve pattern on a storage device, the curve data usable by a camera controller; tracking the movement of the series of curves of the adjusted, predefined facial curve pattern on the performer'"'"'s face during a motion capture session by performing the operations of; capturing sequences of real-time video images of the adjusted, predefined facial curve pattern as the performer makes facial movements during a performance; and correlating each of the captured real-time video images with the curve data read by one or more of the camera controllers to uniquely identify each of the curves and intersection points between the curves; generating motion data representing movement of each of the curves and the intersection points between the curves across the sequences of real-time video images; and storing the motion data on a storage device, the motion data usable for reconstructing the facial movements of the performer during the performance; wherein a first set of one or more cameras capture a first sequence of real-time video images of the adjusted, predefined facial curve pattern from a first orientation as the performer makes facial movements during a performance and a second set of one or more cameras capture a second sequence of real-time video images of the adjusted, predefined facial curve pattern from a second orientation as the performer makes facial movements during a performance; wherein generating the predefined facial curve pattern comprises defining a plurality of curve types and categorizing each of the uniquely identifiable curves as one of the defined curve types, the curve types including contours, radials, and transition curves; wherein the curve data comprises;
a list of curves organized by type, with each curve having a unique identifier associated therewith;
for each curve, an ordered list of other curves that it intersects along its length; and
a list of surface patches, each defined by the curves that make up its sides.
-
-
21. A method for performing motion capture using curves comprising:
-
specifying a predefined facial curve pattern comprising a set of intersecting curves, wherein each individual curve in the set is uniquely identifiable; adjusting the predefined facial curve pattern based on a topology of a performer'"'"'s face and applying the adjusted, predefined facial curve pattern to the performer'"'"'s face; storing the adjusted, predefined facial curve pattern on a storage device, the curve data usable by a camera controller; tracking the movement of the series of curves of the adjusted, predefined facial curve pattern on the performer'"'"'s face during a motion capture session by performing the operations of; capturing sequences of real-time video images of the adjusted, predefined facial curve pattern as the performer makes facial movements during a performance; and correlating each of the captured real-time video images with the curve data read by one or more of the camera controllers to uniquely identify each of the curves and intersection points between the curves; generating motion data representing movement of each of the curves and the intersection points between the curves across the sequences of real-time video images; and storing the motion data on a storage device, the motion data usable for reconstructing the facial movements of the performer during the performance; wherein a first set of one or more cameras capture a first sequence of real-time video images of the adjusted, predefined facial curve pattern from a first orientation as the performer makes facial movements during a performance and a second set of one or more cameras capture a second sequence of real-time video images of the adjusted, predefined facial curve pattern from a second orientation as the performer makes facial movements during a performance; wherein generating the predefined facial curve pattern comprises defining a plurality of curve types and categorizing each of the uniquely identifiable curves as one of the defined curve types, the curve types including contours, radials, and transition curves; quantizing each of the curves into a plurality of line segments between each of the intersection points; and defining a plurality of triangles using the endpoints of each of the line segments and the intersection points of the curves as vertices of the triangles.
-
Specification