Verizon Patent and Licensing Inc.

Solmaz Hajmohammadi

Daniel Kopeinigg

Sourabh Khire

Arthur A Van Hoff

Denny Breitenfeld

Simion Venshtain

Philip Lee

Andrew Walkingshaw

Garry Guan

Richard A Paris

Christopher Redmann

Ching-Kit Chan

Charles LePere

Trevor Howarth

Daniel M Michler

Jie Yu

An illustrative camera calibration system may identify a common feature present within a scene and visible in each of 1) a first image of the scene captured by a first camera from a first pose represented by a first set of calibration parameters, and 2) a second image of the scene captured by a second camera from a second pose represented by a second set of calibration parameters. The system may estimate a 3D position of the common feature with respect to a global coordinate system associated with the scene and determine that the first set of calibration parameters is more accurate than the second set of calibration parameters. In response to this determination, the system may revise the second set of calibration parameters based on the estimated 3D position of the common feature and the first set of calibration parameters. Corresponding methods and systems are also disclosed.

Calibration methods and systems for an under-calibrated camera capturing a scene

An illustrative 3D modeling system generates a hair orientation image based on a 2D image depicting a human subject having hair. The hair orientation image includes semantic segmentation data indicating whether image elements of the hair orientation image depict the hair of the human subject or content other than the hair. The hair orientation image further includes hair flow data indicating a flow direction of the hair depicted by various image elements. Based on the hair orientation image, the 3D modeling system generates a 3D model of the hair of the human subject. The 3D modeling system then integrates the 3D model of the hair of the human subject with a 3D model of a body of the human subject. Corresponding methods and systems are also disclosed.

Methods and systems for 3D modeling of a human subject having hair based on 2D imagery

An illustrative 3D modeling system accesses 1) a 3D body model bounded by vertices forming an unadorned body (without hair or clothing), and 2) a 2D image depicting an adorned subject (with hair and/or clothing). The set of vertices is configurable to simulate different subjects based on a set of parameters and the adorned subject is outlined in the 2D image by a set of silhouette pixels. The 3D modeling system maps the silhouette pixels to vertices of a particular cross section of the 3D body model using an optimization function configured to even out a distribution of silhouette pixels to vertices. Based on the mapping, the 3D modeling system defines the set of parameters to deform the 3D body model such that the particular cross section conforms to an outline of the adorned subject formed by the set of silhouette pixels. Corresponding methods and systems are also disclosed.

Methods and systems for deforming a 3D body model based on a 2D image of an adorned subject

An illustrative multiscale data system determines a first distance between a first object in a scene and a virtual vantage point at the scene. The multiscale data system also determines a second distance between a second object in the scene and the virtual vantage point. In an example in which the second distance is greater than the first distance, the multiscale data system generates, based on the first and second distances, a tiled representation associated with the virtual vantage point. The tiled representation in this example includes a first representation of the first object rendered at a first quality level and a second representation of the second object rendered at a second quality level lower than the first quality level. Corresponding methods and systems are also disclosed.

Methods and systems for data representing objects at different distances from a virtual vantage point

An illustrative volumetric capture system accesses a machine learning model associated with bodies of a particular body type, as well as a two-dimensional (2D) image captured by a capture device located at a real-world scene. The 2D image depicts a body of the particular body type that is present at the real-world scene. Using the machine learning model and based on the 2D image, the volumetric capture system identifies a 2D joint location, from a perspective of the capture device, of a particular joint of the body. The volumetric capture system also generates a three-dimensional (3D) reference model of the body that represents the particular joint of the body at a 3D joint location that is determined based on the 2D joint location identified using the machine learning model. Corresponding methods and systems are also disclosed.

Methods and systems for applying machine learning to volumetric capture of a body in a real-world scene

An exemplary three-dimensional (3D) simulation system accesses a two-dimensional (2D) video image captured by a video capture device and that depicts a bounded real-world scene and a real-world object present within the bounded real-world scene. The 3D simulation system accesses respective 3D models of the bounded real-world scene and the real-world object. Based on the 2D video image, the 3D simulation system tracks a spatial characteristic of the real-world object relative to the bounded real-world scene. Based on the tracked spatial characteristic of the real-world object and the 3D models of the bounded real-world scene and the real-world object, the 3D simulation system generates a 3D simulation of the bounded real-world scene within which the real-world object is simulated in accordance with the tracked spatial characteristic of the real-world object. Corresponding methods and systems are also disclosed.

Methods and systems for performing 3D simulation based on a 2D video image

An illustrative volumetric capture system accesses a two-dimensional ('2D') image captured by a capture device and depicting a first subject of a particular subject type. The volumetric capture system generates a custom three-dimensional ('3D') model of the first subject by identifying a parameter representative of a characteristic of the first subject, applying the parameter to a parametric 3D model to generate a custom mesh, and applying a custom texture based on the 2D image to the custom mesh. The volumetric capture system also accesses a motion capture video depicting motion performed by a second subject of the particular subject type. Based on the motion capture video, the volumetric capture system animates the custom 3D model of the first subject to cause the custom 3D model to perform the motion performed by the second subject. Corresponding methods and systems are also disclosed.

Methods and systems for generating an animated 3D model based on a 2D image

A method includes receiving two-dimensional video streams from a plurality of cameras, the two-dimensional video streams including multiple angles of a sporting event. The method further includes determining boundaries of the sporting event from the two-dimensional video streams. The method further includes identifying a location of a sporting object during the sporting event. The method further includes identifying one or more players in the sporting event. The method further includes identifying poses of each of the one or more players during the sporting event. The method further includes generating a three-dimensional model of the sporting event based on the boundaries of the sporting event, the location of the sporting object during the sporting event, and the poses of each of the one or more players during the sporting event. The method further includes generating a simulation of the three-dimensional model.

Generating three-dimensional content from two-dimensional images

An illustrative multiscale data system obtains first and second datasets representative of respective first and second objects in a scene as viewed from a virtual vantage point at the scene. The system determines respective distances between the respective first and second objects and the virtual vantage point. Based on these distances, the system scales one or more respective representations of the objects rendered from the datasets to cause a first representation of the first object to be of a first quality level higher than a second quality level of a second representation of the second object. The system also generates a tiled representation that is associated with the virtual vantage point and includes the first representation of the first object at the first quality level and the second representation of the second object at the second quality level. Corresponding methods and systems are also disclosed.

Growing community of inventors

Palo Alto, CA, United States of America

Solmaz Hajmohammadi