US 12,014,514 B2Granted

Target classification system

AssigneeMorphix, Inc.

InventorJonathan Lovegrove

Life of this patent

Estimating the term…

Loading patent…

Abstract

One example provides a target classification system comprising a display subsystem configured to display an image captured by a camera of an in-field device. The image includes one or more targets. The target classification system is configured to receive a user input indicating a location of the one or more targets in a screen space coordinate system of the display subsystem. Location information in a world space coordinate system is determined by receiving a pose of the camera; using the pose of the camera and the location in the screen space to trace a ray; and using at least a position of the camera and an orientation of the ray to generate coordinates in the world space. Target classification information is determined, and targeting data is output comprising the coordinates in the world space and the target classification information.

Claims20

Dependent claims

A target classification system comprising:

a display subsystem configured to display an image captured by a camera of a first in-field device, the image including one or more targets;
a user input device configured to receive user input corresponding to locations in the image displayed on the display subsystem;
a processor;
and
a memory storing instructions executable by the processor to,
receive a user input from the user input device indicating a location of the one or more targets in a screen space coordinate system of the display subsystem,
determine location information for the one or more targets in a world space coordinate system of the first in-field device by:
receiving, from a pose sensor of the first in-field device, a position of the camera and a bearing from the camera to the one or more targets,
using the position of the camera, the bearing, and the location of the one or more targets in the screen space to trace a ray between the camera and the one or more targets in the world space, wherein the ray comprises a first line of sight from the first in-field device to the one or more targets in the world space,
receiving at least a second line of sight from at least a second in-field device,
determining one or more intersection points where the ray intersects with at least the second line of sight,
generating an input vector comprising the one or more intersection points, and
inputting the input vector into a trained target location model to thereby cause the trained target location model to output coordinates of the one or more targets in the world space,
determine target classification information for the one or more targets by:
tagging the one or more targets with a first target classification when the user input indicates a first input type, and
tagging the one or more targets with a second target classification when the user input indicates a second input type, and
outputting targeting data comprising the coordinates of the one or more targets in the world space and the target classification information.

14 more elements

At a computing device, a method for classifying a target, the method comprising:

displaying, via a display subsystem, an image captured by a camera of a first in-field device, the image including one or more targets;
receiving a user input from a user input device indicating a location of the one or more targets in a screen space coordinate system of the display subsystem;
determining location information for the one or more targets in a world space coordinate system of the first in-field device by,
receiving, from a pose sensor of the first in-field device, a position of the camera and a bearing from the camera to the one or more targets,
using the position of the camera, the bearing, and the location of the one or more targets in the screen space to trace a ray between the camera and the one or more targets in the world space, wherein the ray comprises a first line of sight from the first in-field device to the one or more targets in the world space,
receiving at least a second line of sight from at least a second in-field device,
determining one or more intersection points where the ray intersects with at least the second line of sight,
generating an input vector comprising the one or more intersection points, and
inputting the input vector into a trained target location model to thereby cause the trained target location model to output coordinates of the one or more targets in the world space;
determining target classification information for the one or more targets by:
tagging the one or more targets with a first target classification when the user input indicates a first input type, and
tagging the one or more targets with a second target classification when the user input indicates a second input type;
and
outputting targeting data comprising the coordinates of the one or more targets in the world space and the target classification information.

11 more elements

A target classification system comprising:

a display subsystem configured to display a graphical user interface (GUI) comprising,
an image captured by a camera of a first in-field device, the image including one or more targets,
a first target classification element configured to receive a user selection indicating a first target classification, and
a second target classification element configured to receive a user selection indicating a second target classification;
a user input device;
a processor;
and
a memory storing instructions executable by the processor to,
receive, from the user input device, a user selection of at least a portion of the image, wherein the user selection comprises at least one pixel in the image, and coordinates of the at least one pixel in the image correspond to a location of the one or more targets in a screen space coordinate system of the display subsystem,
determine location information for the one or more targets in a world space coordinate system of the first in-field device by,
receiving, from a pose sensor of the first in-field device, a position of the camera and a bearing from the camera to the one or more targets,
using the position of the camera, the bearing, and the location of the one or more targets in the screen space to trace a ray between the camera and the one or more targets in the world space, wherein the ray comprises a first line of sight from the first in-field device to the one or more targets in the world space,
receiving at least a second line of sight from at least a second in-field device,
determining one or more intersection points where the ray intersects with at least the second line of sight,
generating an input vector comprising the one or more intersection points,
inputting the input vector into a trained target location model to thereby cause the trained target location model to output coordinates of the one or more targets in the world space,
determine target classification information for the one or more targets by:
tagging the one or more targets with the first target classification when the first target classification element is selected, and
tagging the one or more targets with the second target classification when the second target classification element is selected, and
outputting targeting data comprising the coordinates of the one or more targets in the world space and the target classification information.

17 more elements

Dependencies are inferred from the claim text.

Facts

Application: 17/647,309Filed 6 Jan 2022 · continuation in part
Priority: 4 Jun 2021
Published: US 2022/0391629 A18 Dec 2022
Granted: 18 Jun 2024
Term ends: Not estimatedThe source records don’t support an estimate.

Classification15 codes

F41G 3/02
Aiming or laying means: using an independent line of sight
G06T 7/70
Image analysis: determining position or orientation of objects or cameras
F41G 3/04
Aiming or laying means: for dispersing fire from a battery ; for controlling spread of shots; for coordinating fire from spaced weapons
F41G 3/165
Sighting devices adapted for indirect laying of fire: using a TV-monitor

Family

Application17/338,885Parent of this continuation-in-part · filed 4 Jun 2021
US 2022/0392098 A1Pre-grant publication · published 8 Dec 2022
US 12,026,235 B2Related patent
US 2022/0391629 A1Published application
US 12,014,514 B2This patent

From the US filing history. Foreign family members aren’t listed.