Optimization of label placements in street level images
First Claim
Patent Images
1. A computer-implemented method for labeling an image, comprising:
- receiving, with one or more computing devices, a two-dimensional, street-level image and a three-dimensional model representing at least a portion of the image'"'"'s content;
identifying, with the one or more computing devices, one or more objects in the two-dimensional, street level image;
determining, with the one or more computing devices, object data for each of the one or more identified objects based at least in part on the three-dimensional model, wherein the object data includes an object type chosen from a set of three or more predetermined types, an object location, and an area occupied by the object;
receiving, with the one or more computing devices, label data representing a label to annotate at least a portion of a given one of the one or more objects in the two-dimensional, street-level image; and
using the determined object data to determine, with the one or more computing devices, a label position for presentation of the label within the two-dimensional, street-level image, such that the label does not obscure presentation of the given object in the two-dimensional, street-level image, wherein the object type is used to determine whether the label can be associated with the given object without obscuring the presentation.
2 Assignments
0 Petitions
Accused Products
Abstract
A system, computer-implemented method and computer-readable medium for labeling an image. A two-dimensional street-level image and a three-dimensional model representing at least a portion of the image'"'"'s content are received. Data representing a label to annotate at least a portion of a building in the two dimensional image is also received. A determination is made using the three dimensional mode as to where position the label for presentation within the two-dimensional model such that the label does not obscure presentation of the building in the two-dimensional image.
-
Citations
23 Claims
-
1. A computer-implemented method for labeling an image, comprising:
-
receiving, with one or more computing devices, a two-dimensional, street-level image and a three-dimensional model representing at least a portion of the image'"'"'s content; identifying, with the one or more computing devices, one or more objects in the two-dimensional, street level image; determining, with the one or more computing devices, object data for each of the one or more identified objects based at least in part on the three-dimensional model, wherein the object data includes an object type chosen from a set of three or more predetermined types, an object location, and an area occupied by the object; receiving, with the one or more computing devices, label data representing a label to annotate at least a portion of a given one of the one or more objects in the two-dimensional, street-level image; and using the determined object data to determine, with the one or more computing devices, a label position for presentation of the label within the two-dimensional, street-level image, such that the label does not obscure presentation of the given object in the two-dimensional, street-level image, wherein the object type is used to determine whether the label can be associated with the given object without obscuring the presentation. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10)
-
-
11. A system for labeling an image, comprising:
-
one or more memories storing a two-dimensional, street-level image and a three-dimensional model representing at least a portion of the image'"'"'s content; one or more processors in communication with the one or more memories and configured to; identify one or more objects in the two-dimensional, street-level image; determine object data for each of the one or more identified objects based at least in part on the three-dimensional model, wherein the object data includes an object type chosen from a set of three or more predetermined types, an object location, and an area occupied by the object; receive label data representing a label to annotate at least a portion of a given one of the identified one or more objects in the two-dimensional, street-level image; and using the determined object data to determine a label position for presentation of the label within the two-dimensional, street-level image, such that the label does not obscure presentation of the given object in the two-dimensional, street-level image, wherein the object type is used to determine whether the label can be associated with the given object without obscuring the presentation. - View Dependent Claims (12, 13, 14, 15, 16, 17)
-
-
18. An article of manufacture including a non-transitory computer-readable medium having instructions retrieved thereon that, when executed by one or more processors, cause the one or more processors to perform operations for labeling an image, comprising:
-
receiving a two-dimensional, street-level image and a three-dimensional model representing at least a portion of the image'"'"'s content; identifying one or more objects in the two-dimensional, street-level image; determining object data for each of the one or more identified objects based at least in part on the three-dimensional model, wherein the object data includes an object type chosen from a set of three or more predetermined types, an object location, and an area occupied by the object; receiving label data representing a label to annotate at least a portion of a given one of the one or more objects in the two-dimensional, street-level image; and using the determined object data to determine a label position for presentation of the label within the two-dimensional, street-level image, such that the label does not obscure presentation of the given object in the two-dimensional, street-level image, wherein the object type is used to determine whether the label can be associated with the given object without obscuring the presentation. - View Dependent Claims (19, 20, 21)
-
-
22. A computer-implemented method for labeling an image, comprising:
-
receiving, with one or more computing devices, a two-dimensional, street-level image and a three-dimensional model representing at least a portion of the image'"'"'s content, wherein the three-dimensional model is generated, in part, by using a search and optimization technique comprising; selecting a shape from a plurality of primitive shapes; and defining geometric parameters of the shape to fit a portion of a point cloud representation of dimensions of one or more objects in the two-dimensional, street level image; identifying, with the one or more computing devices, one or more objects in the two-dimensional, street level image; determining, with the one or more computing devices, object data for each of the one or more identified objects based at least in part on the three-dimensional model, wherein the object data includes an object type, an object location, and an area occupied by the object; receiving, with the one or more computing devices, label data representing a label to annotate at least a portion of a given one of the one or more objects in the two-dimensional, street-level image; and determining, with the one or more computing devices and based on the determined object data, a label position for presentation of the label within the two-dimensional, street-level image, such that the label does not obscure presentation of the given object in the two-dimensional, street-level image, wherein the object type is used to determine whether the label can be associated with the given object without obscuring the presentation. - View Dependent Claims (23)
-
Specification