Camera assessment techniques for autonomous vehicles
AssigneeWaymo LLC
InventorsCarl Warren Craddock, Andreas Wendel
Life of this patent
Estimating the term…
Abstract
The disclosure relates to assessing operation of two or more cameras. These cameras may be a group of cameras of a perception system of a vehicle having an autonomous driving mode. A first image captured by a first camera and a second image captured by a second camera may be received. A first feature vector for the first image and a second feature vector for the second image may be generated. A similarity score may be determined using the first feature vector and the second feature vector. This similarity score may be used to assess the operation of the two cameras and an appropriate action may be taken.
Claims20
A method for assessing camera operation to make driving decisions of a vehicle in an autonomous driving mode, the method comprising:
- receiving, by one or more processors, first images captured by a first camera;
- receiving, by the one or more processors, second images captured by a second camera, the first camera and the second camera having an overlapping field of view;
- cropping, by the one or more processors, each of the first images and the second images to include only portions of the first images and the second images that correspond to the overlapping field of view;
- reducing, by the one or more processors, a size of each of the cropped first images and the cropped second images, wherein reducing the size of each of the cropped first images and the cropped second images includes thumbnailing the cropped first images and the cropped second images;
- generating, by the one or more processors, first feature vectors for the reduced first images and second feature vectors for the reduced second images;
- comparing, by the one or more processors, the first feature vectors to the second feature vectors to generate similarity scores, wherein each similarity score corresponds to a pair of feature vectors including at least one of the first feature vectors and at least one of the second feature vectors;
- comparing, by the one or more processors, pairs of the similarity scores over time in order to assess operation of the first camera and the second camera;
- generating, by the one or more processors, a driving decision for the vehicle based on whether the assessment indicates an abrupt change in the compared pairs of the similarity scores;
- and
- controlling, by the one or more processors, the vehicle to pull over or stop based on the driving decision.
A method for assessing camera operation to make driving decisions of a vehicle in an autonomous driving mode, the method comprising:
- receiving, by one or more processors, first images captured by a first camera;
- receiving, by the one or more processors, second images captured by a second camera, the first camera and the second camera having an overlapping field of view;
- cropping, by the one or more processors, each of the first images and the second images to include only portions of the first images and the second images that correspond to the overlapping field of view;
- reducing, by the one or more processors, a size of each of the cropped first images and the cropped second images, wherein reducing the size of each of the cropped first images and the cropped second images includes thumbnailing the cropped first images and the cropped second images;
- assessing, by the one or more processors, operation of the first camera and the second camera based on the reduced first images and the reduced second images;
- generating, by the one or more processors, a driving decision for the vehicle based on the assessing;
- and
- controlling, by the one or more processors, the vehicle to pull over or stop based on the driving decision.
Dependencies are inferred from the claim text.