Urban environment labelling
AssigneeBlue Vision Labs UK Limited
InventorsPeter Ondruska, Lukas Platinsky, Giacomo Dabisias
Life of this patent
Estimating the term…
Abstract
The present invention relates to a method and system for automatic localisation of static objects in an urban environment. More particularly, the present invention relates to the use of noisy 2-Dimensional (2D) image data to identify and determine 3-Dimensional (3D) positions of objects in large scale urban or city environments. Aspects and/or embodiments seek to provide a method, system, and vehicle for automatically locating static 3D objects in urban environments by using a voting-based triangulation technique. Aspects and/or embodiments also provide a method for updating map data after automatically new 3D static objects in an environment.
Claims20
A computer-implemented method comprising:
- determining, by a computing system, a data set including images associated with an object captured in an area;
- distributing, by the computing system, the data set into at least a first cluster that is associated with the object and a second cluster that is associated with the object, wherein the first and second clusters are based on an identification of the images included in the data set;
- and
- determining, by the computing system, a position of the object based on triangulating the object according to at least the first cluster and the second cluster, wherein the determining the position of the object comprises:
- performing, by the computing system, voting-based triangulation on the first cluster, wherein a vote is based on a distance between an estimated position and a position of a detection of the object in an image;
- and
- performing, by the computing system, the voting-based triangulation on the second cluster.
A system comprising:
- at least one processor;
- and
- a memory storing instructions that, when executed by the at least one processor, cause the system to perform:
- determining a data set including images associated with an object captured in an area;
- distributing the data set into at least a first cluster that is associated with the object and a second cluster that is associated with the object, wherein the first and second clusters are based on an identification of the images included in the data set;
- and
- determining a position of the object based on triangulating the object according to at least the first cluster and the second cluster, wherein the determining the position of the object comprises:
- performing voting-based triangulation on the first cluster, wherein a vote is based on a distance between an estimated position and a position of a detection of the object in an image;
- and
- performing the voting-based triangulation on the second cluster.
A non-transitory computer-readable storage medium including instructions that, when executed by at least on processor of a computing system, cause the computing system to perform:
- determining a data set including images associated with an object captured in an area;
- distributing the data set into at least a first cluster that is associated with the object and a second cluster that is associated with the object, wherein the first and second clusters are based on an identification of the images included in the data set;
- and
- determining a position of the object based on triangulating the object according to at least the first cluster and the second cluster, wherein the determining the position of the object comprises:
- performing voting-based triangulation on the first cluster, wherein a vote is based on a distance between an estimated position and a position of a detection of the object in an image;
- and
- performing the voting-based triangulation on the second cluster.
Dependencies are inferred from the claim text.