US 12,409,859 B1
Object detection using multispectral data
Philippe Martin Burlina, Rockville, MD (US); Subhasis Das, San Mateo, CA (US); and Xinyu Xu, San Jose, CA (US)
Assigned to Zoox, Inc., Foster City, CA (US)
Filed by Zoox, Inc.
Filed on Mar. 17, 2023, as Appl. No. 18/123,165.
Int. Cl. G06V 20/58 (2022.01); B60W 30/09 (2012.01); B60W 60/00 (2020.01); G06V 10/143 (2022.01); G06V 10/58 (2022.01); G06V 10/74 (2022.01); G06V 10/764 (2022.01)
CPC B60W 60/0015 (2020.02) [B60W 30/09 (2013.01); G06V 10/143 (2022.01); G06V 10/58 (2022.01); G06V 10/761 (2022.01); G06V 10/764 (2022.01); G06V 20/58 (2022.01); B60W 2420/40 (2013.01); B60W 2554/40 (2020.02); B60W 2720/10 (2013.01)] 20 Claims
OG exemplary drawing
 
1. An autonomous vehicle comprising:
a sensor coupled to the autonomous vehicle;
one or more processors; and
one or more non-transitory computer readable media storing instructions executable by the one or more processors, wherein the instructions, when executed, cause the autonomous vehicle to perform operations comprising:
receiving, from the sensor, sensor data comprising data points in a plurality of spectral bands of the sensor, the plurality of spectral bands including visible and non-visible wavelengths;
determining a subset of the sensor data associated with a candidate region in an environment in which an autonomous vehicle is disposed;
determining, based on applying dimensionality reduction to the subset of the sensor data, a reduced data;
generating, based on the reduced data, a graph representation indicating spatial relationships between spectral responses associated with portions of the candidate region,
wherein nodes of the graph representation are indicative of the spectral responses, and edges of the graph representation are indicative of a spatial proximity between the nodes;
determining, based on a similarity between the graph representation and a known signature associated with an object class, a probability of presence of an object of the object class in the candidate region; and
controlling, based at least in part on the probability associated with the presence of the object, the autonomous vehicle.