Three-dimensional modeling of an object
AssigneeGeneral Electric Company
InventorsEric Michael Gros, Junli Ping, Arpit Jain, Ming-Ching Chang, Peter Henry Tu
Life of this patent
Estimating the term…
Abstract
The present approach relates to the use of a point cloud of an object to initialize or seed a space carving technique used to generate a 3D model of the object. In one implementation, feature matching is performed on 2D images, with matched features constituting the points of a point cloud model. The point cloud generated in this manner, is one input of a foreground/background segmentation algorithm, which generates a set of segmented 2D images used by a space carving routine to generate a 3D model of the object.
Claims20
A system, comprising:
- an autonomous robot configured to acquire image data of an object via at least one sensor, wherein the image data comprises a plurality of images of the object;
- and
- one or more processors communicatively coupled to the autonomous robot and configured to:
- receive the image data from the autonomous robot;
- perform a feature matching operation on the plurality of images to generate a three dimensional (3D) point cloud model of the object;
- subsequent to performing the feature matching operation, segment foreground and background of at least a subset of the plurality of images based on the 3D point cloud model to generate segmented images defining silhouettes of the object;
- and
- perform a space carving operation on an initial volume using the segmented images to generate a 3D model of the object.
A system, comprising:
- a memory storing instructions;
- and
- one or more processors configured to execute the instructions, wherein the instructions, when executed by the one or more processors, cause the one or more processors to:
- instruct an autonomous robot to acquire a plurality of images of an object via at least one sensor of the autonomous robot;
- generate, via a feature matching operation, a point cloud model of the object based on the plurality of images;
- subsequent to performing the feature matching operation, segment at least a subset of the plurality images based on the point cloud model to generate segmented images distinguishing silhouettes of the object from a background surrounding the object;
- and
- perform a space carving operation on an initial volume via projection of the silhouettes on the initial volume to generate a 3D model of the object.
A non-transitory computer readable medium having instructions stored thereon, wherein the instructions, when executed by a processor, cause the processor to:
- receive a plurality of images of an object from at least one sensor;
- perform a feature matching operation on the plurality of images to generate a three-dimensional (3D) point cloud model of the object;
- subsequent to performing the feature matching operation, segment foreground and background of at least a subset of the plurality of images based on the 3D point cloud model to generate segmented images defining silhouettes of the object;
- project the silhouettes on an initial volume to generate a plurality of projection regions;
- and
- perform a space carving operation on the initial volume by excluding volumes outside of the plurality of projection regions from the initial volume to generate a 3D model of the object.
Dependencies are inferred from the claim text.