METHOD AND APPARATUS FOR ESTIMATING BODY SHAPE
First Claim
1. A computer implemented method of estimating a shape of a body of an individual, comprising the steps of:
- obtaining input data including data representing said body in a plurality of poses; and
fitting a parametric body model of said body to the data representation of said body contained in the input data in said plurality of poses to generate multiple sets of pose parameters and at least one set of shape parameters, said at least one set of shape parameters being consistent with said plurality of poses.
3 Assignments
0 Petitions
Accused Products
Abstract
A system and method of estimating the body shape of an individual from input data such as images or range maps. The body may appear in one or more poses captured at different times and a consistent body shape is computed for all poses. The body may appear in minimal tight-fitting clothing or in normal clothing wherein the described method produces an estimate of the body shape under the clothing. Clothed or bare regions of the body are detected via image classification and the fitting method is adapted to treat each region differently. Body shapes are represented parametrically and are matched to other bodies based on shape similarity and other features. Standard measurements are extracted using parametric or non-parametric functions of body shape. The system components support many applications in body scanning, advertising, social networking, collaborative filtering and Internet clothing shopping.
-
Citations
86 Claims
-
1. A computer implemented method of estimating a shape of a body of an individual, comprising the steps of:
-
obtaining input data including data representing said body in a plurality of poses; and fitting a parametric body model of said body to the data representation of said body contained in the input data in said plurality of poses to generate multiple sets of pose parameters and at least one set of shape parameters, said at least one set of shape parameters being consistent with said plurality of poses. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8)
-
-
9. A computer implemented method of estimating a shape of a body of an individual, comprising the steps of:
-
obtaining input data including data representing a body of an individual in at least one pose, said body being at least partially clothed; detecting, using the data representation of said body, at least a clothed portion of the data representation of said body, the clothed portion having an associated looseness characteristic; and obtaining an estimation of body shape under at least the clothed portion of said body by optimizing an objective function based at least in part on the looseness characteristic. - View Dependent Claims (10, 11, 12)
-
-
13. A computer implemented method of estimating a shape of a body of an individual, comprising the steps of:
-
obtaining one or more input images from at least one camera, said one or more input images containing depictions of said body in at least one pose, said at least one camera having an associated set of camera calibration parameters; and obtaining an estimation of the shape of the body of said individual by optimizing an objective function based at least in part on the set of camera calibration parameters. - View Dependent Claims (14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30)
-
-
31. A computer implemented method of estimating a shape of a body of an individual, comprising the steps of:
-
obtaining input data including data representing said body in at least one pose, said body having at least one associated attribute, the at least one attribute of said body being subject to at least one constraint; fitting a parametric body model of said body to the data representation of said body while substantially satisfying said at least one constraint on the at least one attribute of said body, said parametric body model being defined by a plurality of parameters including at least one set of pose parameters and a set of shape parameters, whereby said parametric body model has an associated shape that is representative of the shape of the body of said individual. - View Dependent Claims (32, 33, 34, 35, 36, 37)
-
-
38. A computer implemented method of obtaining camera calibration and image segmentation, comprising the steps of:
-
obtaining, from at least one camera, image data associated with at least one image of an object in front of at least one surface, said at least one surface having a predetermined multi-chroma pattern with predetermined pattern dimensions; in a first processing step, processing the image data from the at least one image of said object to obtain one or more intrinsic or extrinsic parameters associated with the camera; and in a second processing step, utilizing the same image data associated with the at least one image of said object, performing segmentation to segment a foreground image of the object from said at least one surface. - View Dependent Claims (39, 40, 41, 42, 43, 44, 45, 46, 47)
-
-
48. A computer implemented method of estimating a measurement associated with a body of an individual, said body having an associated shape and at least one associated attribute, comprising the steps of:
-
obtaining a mapping of a representation of the shape of the body of said individual to the at least one attribute of said body; obtaining body shape information pertaining to the shape of the body of said individual; and estimating at least one characteristic associated with the body of said individual using the body shape information for said individual and the mapping of said representation of the shape of the body of said individual to the at least one attribute of said body. - View Dependent Claims (49, 50, 51, 52, 53, 54, 55)
-
-
56. A method of matching attributes associated with bodies of a plurality of individuals, comprising the steps of:
-
obtaining input data representing a body of a first individual; generating, using said input data, information defining a parametric body model of said first individual, the parametric body model of said first individual having at least one associated attribute; accessing, from a database, information defining at least one parametric body model of at least one second individual, the parametric body model of the second individual having at least one associated attribute; computing a measure of similarity between the at least one attribute of the parametric body model of said first individual and the at least one attribute of said at least one parametric body model of said at least one second individual; and generating a value indicative of the similarly between the attributes of the parametric body models of the first and second individuals based at least in part on the computed measure of similarity. - View Dependent Claims (57, 58, 59, 60, 61, 62, 63)
-
-
64. A computer implemented method of estimating a shape of a body of an individual, comprising the steps of:
-
obtaining a database containing body shape information, body pose information, and feature information, said body shape information pertaining to shapes of a plurality of bodies of individuals, said body pose information pertaining to poses associated with said plurality of bodies, said feature information comprising at least one transformation of real or simulated sensor data for each of said plurality of bodies; generating, using said body shape information, said body pose information, and said feature information contained in said database, a mapping of feature information to body shape and pose information; obtaining input sensor data representing the body of said individual; obtaining feature information for said individual from said input sensor data, said feature information for said individual comprising at least one transformation of said input sensor data for the body of said individual; and estimating, using the feature information for said individual and said mapping of feature information to body shape and pose information, the shape of the body of said individual. - View Dependent Claims (65, 66, 67)
-
-
68. A computer implemented method of determining an indication of a product associated with a human body, comprising the steps of:
-
obtaining a representation of at least one aspect of a body of an individual; comparing the representation of the at least one aspect of the body of said individual to at least one representation of at least one aspect of each of a plurality of bodies of different individuals contained in a database, wherein each representation associated with each of the different individuals has a product associated therewith; and obtaining a ranking of similarity of the representation of the at least one aspect of the body of said individual to the representation of the at least one aspect of the bodies of said plurality of different individuals. - View Dependent Claims (69, 70, 71, 72, 73, 74)
-
-
75. A system for estimating a shape of a body of an individual, comprising:
-
an input device operative to obtain input data including data representing said body in a plurality of poses; and at least one processor operative to execute at least one program out of at least one memory to fit a parametric body model of said body to the data representation of said body contained in the input data in said plurality of poses to generate multiple sets of pose parameters and at least one set of shape parameters, said at least one set of shape parameters being consistent with said plurality of poses.
-
-
76. A system for estimating a shape of a body of an individual, comprising:
-
an input device operative to obtain input data including data representing a body of an individual in at least one pose, said body being at least partially clothed; and at least one processor operative to execute at least one program out of at least one memory to detect, using the data representation of said body, at least a clothed portion of the data representation of said body, the clothed portion having an associated looseness characteristic, and to generate an estimation of body shape under at least the clothed portion of said body by optimizing an objective function based at least in part on the looseness characteristic.
-
-
77. A system for estimating a shape of a body of an individual, comprising:
-
an input device operative to obtain a plurality of input images from at least one camera, said plurality of input images containing depictions of said body in at least one pose, said at least one camera having an associated set of camera calibration parameters; and at least one processor operative to execute at least one program out of at least one memory to generate an estimation of the shape of the body of said individual by optimizing an objective function based at least in part on the set of camera calibration parameters.
-
-
78. A system for estimating a shape of a body of an individual, comprising:
-
an input device operative to obtain input data including data representing said body in at least one pose, said body having at least one associated attribute, the at least one attribute of said body being subject to at least one constraint; and at least one processor operative to execute at least one program out of at least one memory to fit a parametric body model of said body to the data representation of said body while substantially satisfying said at least one constraint on the at least one attribute of said body, said parametric body model being defined by a plurality of parameters including at least one set of pose parameters and a set of shape parameters, whereby said parametric body model has an associated shape that is representative of the shape of the body of said individual.
-
-
79. A system for obtaining camera calibration and image segmentation, comprising:
-
at least one surface having a predetermined multi-chroma pattern with predetermined pattern dimensions; an input device operative to obtain, from at least one camera, image data associated with at least one image of an object in front of said at least one surface; and at least one processor operative to execute at least one program out of at least one memory; to process the image data from the at least one image of said object to obtain one or more intrinsic or extrinsic parameters associated with the camera; and utilizing the same image data associated with the at least one image of said object, to perform segmentation to segment a foreground image of the object from said at least one surface. - View Dependent Claims (80)
-
-
81. A system for estimating a measurement associated with a body of an individual, said body having an associated shape and at least one associated attribute, comprising:
-
at least one processor operative to execute at least one program out of at least one memory; to receive body shape information pertaining to the shape of the body of said individual; and to estimate at least one characteristic associated with the body of said individual using the body shape information for said individual and a mapping of a representation of the shape of the body of said individual to the at least one attribute of said body.
-
-
82. A system for matching attributes associated with bodies of a plurality of individuals, comprising:
-
at least one processor operative to execute at least one program out of at least one memory; to receive input data representing a body of a first individual; to generate, using said input data, information defining a parametric body model of said first individual, the parametric body model of said first individual having at least one associated attribute; to access, from a database, information defining at least one parametric body model of at least one second individual, the parametric body model of the second individual having at least one associated attribute; to compute a measure of similarity between the at least one attribute of the parametric body model of said first individual and the at least one attribute of said at least one parametric body model of said at least one second individual; and to generate a value indicative of the similarly between the attributes of the parametric body models of the first and second individuals based at least in part on the computed measure of similarity.
-
-
83. A system for estimating a shape of a body of an individual, comprising:
-
a database containing body shape information, body pose information, and feature information, said body shape information pertaining to shapes of a plurality of bodies of individuals, said body pose information pertaining to poses associated with said plurality of bodies, said feature information comprising at least one transformation of sensor data for each of said plurality of bodies; at least one input device operative to obtain input sensor data representing the body of said individual; and at least one processor operative to execute at least one program out of at least one memory; to generate feature information for said individual from said input sensor data, said feature information for said individual comprising at least one transformation of said input sensor data for the body of said individual; to generate, using said body shape information, said body pose information, and said feature information contained in said database, a mapping of feature information to body shape and pose information; and to estimate, using the feature information for said individual and said mapping of feature information to body shape and pose information, the shape of the body of said individual.
-
-
84. A system for determining an indication of a product associated with a human body, comprising:
-
at least one processor operative to execute at least one program out of at least one memory; to receive a representation of at least one aspect of a body of an individual; to compare the representation of the at least one aspect of the body of said individual to at least one representation of at least one aspect of each of a plurality of bodies of different individuals contained in a database, wherein each representation associated with each of the different individuals has a product associated therewith; and to generate a ranking of similarity of the representation of the at least one aspect of the body of said individual to the representation of the at least one aspect of the bodies of said plurality of different individuals.
-
-
85. A computer implemented method of determining attribute similarities, comprising the steps of:
-
receiving input data representing a body of a first individual; generating, using said input data, information defining a parametric body model of said first individual, the parametric body model of said first individual having at least one associated attribute; accessing, from a database, information corresponding to at least one attribute specification; and in a computing step, computing a score indicative of a similarity between the at least one attribute associated with the first individual and each of the at least one attribute specification.
-
-
86. A system for determining attribute similarities, comprising:
-
At least one processor operative to execute at least one program out of at least one memory; to receive input data representing a body of a first individual; to generate, using said input data, information defining a parametric body model of said first individual, the parametric body model of said first individual having at least one associated attribute; to access, from a database, information corresponding to at least one attribute specification; and to compute a score indicative of a similarity between the at least one attribute associated with the first individual and each of the at least one attribute specification.
-
Specification