SIMULATED CONSISTENCY CHECK FOR POINTS OF INTEREST ON THREE-DIMENSIONAL MAPS
The present disclosure describes approaches to camera re-localization that improve the accuracy of re-localization determinations by performing simulated consistency checks for three-dimensional maps. Client devices associated with users of a location-based application transmit image scans to a game server, which divides the received scan data into mapping sets used to generate 3D maps of environments and validation sets used to test the accuracy of the maps. To perform the testing, the game server identifies query scans in the validation set having GPS coordinates within a threshold distance of the mapped location and uses the 3D map of the environment to generate a pose estimate for each frame. The results of the localization queries are analyzed by comparing differences between the localization pose estimates and differences between the poses of independent pairs of frames in the query scan to evaluate the accuracy of the 3D map.
1 . A computer-implemented method comprising:
accessing an image captured by a camera of a client device;
determining a position of the client device based on the image, the position having a confidence score representing a probability that the position is correct;
comparing the confidence score to a confidence threshold; and
initiating an augmented reality experience at the client device responsive to the confidence score meeting the threshold.
2 . The computer-implemented method of claim 1 , wherein the confidence threshold is selected based on a parameter.
3 . The computer-implemented method of claim 2 , wherein the parameter is either speed or accuracy, with the parameter being accuracy setting the confidence threshold higher than the parameter being speed.
4 . The computer-implemented method of claim 1 , wherein the position of the client device comprises a pose of the client device.
5 . The computer-implemented method of claim 1 , wherein determining the position of the client device comprises obtaining a 3D map, the position being determined by a comparison of data derived from the image to the 3D map.
6 . The computer-implemented method of claim 5 , wherein 3D map is associated with a localizability score that represents a likelihood that the position of the client device determined using the 3D map is accurate.
7 . The computer-implemented method of claim 6 , wherein the localizability score is a likelihood that the 3D map will provide an accurate position for the client device to at least a threshold level of accuracy within a specified period of time.
8 . The computer-implemented method of claim 6 , wherein the localizability score is generated based on a second confidence score of a pose estimate for a pair of frames in a query scan and a scan quality score for the query scan.
9 . The computer-implemented method of claim 8 , wherein the second confidence score for the pose estimate represents a probability that the pose estimate is correct and the scan quality score indicates a level of localization and consistency of the query scan.
10 . The computer-implemented method of claim 6 , wherein the localizability score of the 3D map is periodically recalculated.
11 . A non-transitory computer-readable storage medium comprising instructions executable by a computing system, the instructions, when executed, causing the computing system to perform operations including:
accessing an image captured by a camera of a client device;
determining a position of the client device based on the image, the position having a confidence score representing a probability that the position is correct;
comparing the confidence score to a confidence threshold; and
initiating an augmented reality experience at the client device responsive to the confidence score meeting the threshold.
12 . The non-transitory computer-readable medium of claim 11 , wherein the confidence threshold is selected based on a parameter.
13 . The non-transitory computer-readable medium of claim 12 , wherein the parameter is either speed or accuracy, with the parameter being accuracy setting the confidence threshold higher than the parameter being speed.
14 . The non-transitory computer-readable medium of claim 11 , wherein the position of the client device comprises a pose of the client device.
15 . The non-transitory computer-readable medium of claim 11 , wherein determining the position of the client device comprises obtaining a 3D map, the position being determined by a comparison of data derived from the image to the 3D map.
16 . The non-transitory computer-readable medium of claim 15 , wherein 3D map is associated with a localizability score that represents a likelihood that the position of the client device determined using the 3D map is accurate.
17 . The non-transitory computer-readable medium of claim 16 , wherein the localizability score is a likelihood that the 3D map will provide an accurate position for the client device to at least a threshold level of accuracy within a specified period of time.
18 . The non-transitory computer-readable medium of claim 16 , wherein the localizability score is generated based on a second confidence score of a pose estimate for a pair of frames in a query scan and a scan quality score for the query scan.
19 . The non-transitory computer-readable medium of claim 18 , wherein the second confidence score for the pose estimate represents a probability that the pose estimate is correct and the scan quality score indicates a level of localization and consistency of the query scan.
20 . The non-transitory computer-readable medium of claim 16 , wherein the localizability score of the 3D map is periodically recalculated.