US 11,113,873 B1Granted

Modeling articulated objects

AssigneeZoox, Inc.

InventorsMichael Carsten Bosse, Yassen Ivanchev Dobrev, Shiwei Sheng

Life of this patent

Estimating the term…

Loading patent…

Abstract

Navigation systems can identify objects in an environment and generate representations of those objects. A representation of an articulated vehicle can include two segments rotated relative to each other about a pivot, with a first segment corresponding to a first portion of the articulated vehicle and the second segment corresponding to a second portion of the articulated vehicle. The representation can be based on a model fit to points that are derived from sensor data and are associated with the object. In some examples, the model can be fit to the points using an expectation maximization algorithm and can be parameterized using attributes of the first and second segments.

Claims19

Dependent claims

A system comprising:

one or more processors;
non-transitory computer-readable media storing instructions executable by the one or more processors to cause the system to perform operations comprising:
receiving a plurality of points associated with an object;
generating a first model representative of the object, the first model comprising a first rectangle connected to a second rectangle at a pivot, the first rectangle comprising a first end, a second end spaced from the first end by a first length, a first side, and a second side spaced from the first side by a first width, the second rectangle comprising a third end, a fourth end spaced from the third end by a second length, a third side, and a fourth side spaced from the third side by a second width, the pivot being coincident with a first midpoint on the second end and a second midpoint on the third end;
determining, for a point of the plurality of points, a first distance of the point to the first rectangle and a second distance of the point to the second rectangle;
determining, for the point and based at least in part on the first distance and the second distance, a first probability that the point is associated with the first rectangle and a second probability that the point is associated with the second rectangle;
generating, based at least in part on the first probability and the second probability, a second model representative of the object, the second model comprising the first rectangle and the second rectangle, wherein at least one of:
the first rectangle is rotated relative to the second rectangle about the pivot, or
at least one of the first width, the second width, the first length, or the second length is altered relative to the first model;
and
controlling, based at least in part on the second model as a representation of the object in an environment of an autonomous vehicle, the autonomous vehicle to navigate the environment relative to the object.

8 more elements

A method comprising:

receiving a plurality of points;
generating a first model representative of the plurality of points, the first model comprising a first geometric shape and a second geometric shape connected to the first geometric shape at a pivot, the first geometric shape being rotated relative to the second geometric shape by a first angle about the pivot;
determining, for a point of the plurality of points, a first probability of the point being associated with the first geometric shape based at least in part on a first distance of the point from the first geometric shape;
determining a second probability of the point being associated with the second geometric shape based at least in part on a second distance of the point from the second geometric shape;
generating, based at least in part on the first probability and the second probability, a second model representative of the plurality of points, the second model comprising the first geometric shape rotated relative to the second geometric shape by a second angle about the point;
and
controlling, based at least in part on the second model as a representation of an object in an environment an autonomous vehicle relative to the object.

4 more elements

An autonomous vehicle comprising:

a sensor configured to generate sensor data;
one or more processors;
and
memory storing processor-executable instructions that, when executed by the one or more processors, configure the vehicle to perform actions comprising:
receiving the sensor data, the sensor data comprising a plurality of points associated with an object in an environment of the vehicle;
generating a first model representative of the plurality of points, the first model comprising a first geometric shape and a second geometric shape connected to the first geometric shape at a pivot, the first geometric shape being rotated relative to the second geometric shape by a first angle about the pivot;
determining, for a point of the plurality of points, a first distance of the point from the first geometric shape;
determining a second distance of the point from the second geometric shape;
determining, based at least in part on the first distance, a first probability of the point being associated with the first geometric shape;
determining, based at least in part on the second distance a second probability of the point being associated with the second geometric shape;
generating, based at least in part on the first probability and the second probability, a second model representative of the plurality of points, the second model comprising the first geometric shape rotated relative to the second geometric shape by a second angle about the point;
and
controlling, based at least in part on the second model, the autonomous vehicle to travel through the environment.

10 more elements

Dependencies are inferred from the claim text.

Facts

Application: 16/586,455Filed 27 Sep 2019
Granted: 7 Sep 2021
Term ends: Not estimatedThe source records don’t support an estimate.

Classification13 codes

G06T 19/20
Manipulating three-dimensional [3D] models or images for computer graphics: editing of three-dimensional [3D] images, e.g. changing shapes or…
G05D 1/0088
Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots: characterized by the…
G06T 7/10
Image analysis: segmentation; Edge detection
G06T 17/00
Three-dimensional [3d] modelling for computer graphics

Family

No related US applications are recorded in the indexed filing history.

Timeline

Citations

Open elsewhere

Building an agent or pulling many patents? Use the API or MCP server instead of scraping pages. Get an API key