Real-Time Bicyclist Detection with Synthetic Training Data
AssigneeHonda Motor Co., Ltd.
InventorBernd Heisele
Life of this patent
Estimating the term…
Abstract
A determination is made in real-time regarding whether a bicyclist is present in a target image. A target image is received. The target image is classified and an error value for the target image is determined using a linear classifier. If the error value does not exceed the threshold value, the classification is outputted. Otherwise, if the error value exceeds the threshold value, the target image is classified using a non-linear classifier.
Claims20
A method for determining in real-time whether a bicyclist is present in a target image, comprising steps of:
- receiving the target image;
- determining a first classification and an error value for the target image using a linear classifier, wherein the error value indicates a probability that the first classification is incorrect;
- responsive to the error value not exceeding a threshold:
- outputting the first classification;
- and
- responsive to the error value exceeding the threshold:
- determining a second classification for the target image using a non-linear classifier;
- and
- outputting the second classification.
A non-transitory computer-readable storage medium configured to store instructions for determining in real-time whether a bicyclist is present in a target image, the instructions when executed by a processor cause the processor to:
- receive the target image;
- determine a first classification and an error value for the target image using a linear classifier, wherein the error value indicates a probability that the first classification is incorrect;
- responsive to the error value not exceeding a threshold:
- output the first classification;
- and
- responsive to the error value exceeding the threshold:
- determine a second classification for the target image using a non-linear classifier;
- and
- output the second classification.
A system for determining in real-time whether a bicyclist is present in a target image, comprising:
- a processor;
- and
- a non-transitory computer-readable storage medium storing instructions, the instructions when executed by the processor cause the processor to:
- receive the target image;
- determine a first classification and an error value for the target image using a linear classifier, wherein the error value indicates a probability that the first classification is incorrect;
- responsive to the error value not exceeding a threshold:
- output the first classification;
- and
- responsive to the error value exceeding the threshold:
- determine a second classification for the target image using a non-linear classifier;
- and
- output the second classification.
Dependencies are inferred from the claim text.