US 11,957,978 B2 · Refining camera re-localization determination using prior pose model

Abstract

The present disclosure describes approaches to camera re-localization that improve the speed and accuracy with which pose estimates are generated by fusing output of a computer vision algorithm with data from a prior model of a geographic area in which a user is located. For each candidate pose estimate output by the algorithm, a game server maps the estimate to a position on the prior model (e.g., a specific cell on a heatmap-style histogram) and retrieves a probability corresponding to the mapped position. A data fusion module fuses, for each candidate pose estimate, a confidence score generated by the computer vision algorithm with the location probability from the prior model to generate an updated confidence score. If an updated confidence score meets or exceeds a score threshold, a re-localization module initiates a location-based application (e.g., a parallel reality game) based on the associated candidate pose estimate.

Claims20

Dependent claims

A computer-implemented method comprising:

receiving camera frame and sensor data from a client device associated with a location-based application;
inputting the received data into a computer vision algorithm to generate candidate pose estimates, each pose estimate representing a prediction of a three-dimensional position and orientation of the client device and having an associated confidence score;
for each candidate pose estimate:
mapping the candidate pose estimate to a position in a prior model of a geographic area in which the client device is located;
determining, using the prior model, a probability that the client device is located at the three-dimensional position;
and
fusing the confidence score with the probability to generate an updated confidence score;
selecting an authoritative candidate pose estimate from the pose estimates using the updated confidence scores;
and
causing the client device to provide location-based content within the location-based application using the authoritative pose estimate.

7 more elements

A non-transitory computer-readable storage medium comprising instructions executable by a processor, the instructions executable to perform operations comprising:

receiving camera frame and sensor data from a client device associated with a location-based application;
inputting the received data into a computer vision algorithm to generate candidate pose estimates, each pose estimate representing a prediction of a three-dimensional position and orientation of the client device and having an associated confidence score;
for each candidate pose estimate:
mapping the candidate pose estimate to a position in a prior model of a geographic area in which the client device is located;
determining, using the prior model, a probability that the client device is located at the three-dimensional position;
fusing the confidence score with the probability to generate an updated confidence score;
selecting an authoritative candidate pose estimate from the pose estimates using the updated confidence scores;
and
causing the client device to provide location-based content within the location-based application using the authoritative pose estimate.

6 more elements

Dependencies are inferred from the claim text.

Refining camera re-localization determination using prior pose model

Life of this patent

Refining camera re-localization determination using prior pose model

Life of this patent

Abstract

Claims20