Method and sensor apparatus for generating an object classification for an object

1. A method of generating an object classification for an object localized in first and second images of objects captured using a light imaging sensor and a radar sensor respectively, the method comprising:

generating image data for the object localized in the first image of the objects captured using the light imaging sensor,

operating an image-based object classifier on the image data to generate an image-based object classification of the object localized in the first image of the objects;

generating radar data corresponding to an echo reflected from the object localized in the second image of the objects captured using the radar sensor,

operating a radar-based object classifier on the radar data to generate a radar-based object classification of the object localized in the second image of the objects,

selecting between the image-based object classification and the radar-based object classification to output as the object classification for the object;

determining if a training condition is met by the radar-based object classification, and

training the radar-based object classifier using the image-based object classification if the training condition is met by the radar-based object classification;

wherein the object is one of a human, animal, or vehicle.

2. The method of claim 1 , wherein the selecting step comprises selecting the radar-based object classification when a radar condition is met.

3. The method of claim 2 , wherein the radar condition is met when a probability value of the image-based object classification is below a threshold.

4. The method of claim 3 , wherein the radar condition is only met when a probability value of the radar-based object is above a threshold.

5. The method of claim 2 , wherein the radar condition is met when the image sensor is at least partially obstructed.

6. The method of claim 2 , wherein the radar condition is met when a low light condition is determined using the image sensor.

7. The method of claim 2 , wherein the radar condition is met when a low contrast condition is determined using the image sensor.

8. The method of claim 2 , wherein the radar condition is met when the image sensor is determined to have an output having a low signal to noise ratio.

9. The method of claim 1 , wherein the selecting step comprises selecting between the image-based object classification and the radar-based object classification in dependence on one or more of a time of day, a light level, a visibility distance, a humidity level, and a weather condition.

10. The method of claim 9 , wherein the time of day, a light level, a visibility distance, a humidity level, and/or weather condition is determined by an independent sensor.

11. The method of claim 1 , wherein the radar-based object classification comprises a radar object probability value.

12. The method of claim 11 , wherein the training condition is met when the radar object probability value is below a first threshold.

13. The method of claim 12 , wherein the training condition is only met when a probability value of the image-based object classification is above a second threshold.

14. The method of claim 1 , wherein the training condition is met when the image-based object classification and the radar-based object classification are substantially mismatched.

15. An apparatus for generating an object classification for an object of a plurality of objects comprising:

a light imaging sensor;

a radar sensor; and

a processor, the processor configured to:

receive image data corresponding to the object localized by a bounding box in a first image of the objects captured by the light imaging sensor;

operate an image-based object classifier on the image data to generate an image-based object classification;

receive radar data corresponding to an echo reflected from the object in a second image of the objects captured by the radar sensor;

operate a radar-based object classifier on the radar data to generate a radar-based object classification;

select between the image-based object classification and the radar-based-object classification to output as the object classification for the object;

determine if a training condition is met by the radar-based object classification; and

train the radar-based object classifier using the image-based object classification when the training condition is met by the radar-based object classification;

wherein the object is at least one of a human, animal, or vehicle.

16. The method of claim 1 wherein:

the image-based object classifier is a neural network based classifier; and

the radar-based object classifier is a neural network based classifier.

17. The method of claim 16 wherein the selecting between the image-based object classification and the radar-based object classification to output as the object classification for the object is performed automatically.

18. The apparatus of claim 15 wherein:

the image-based object classifier is a neural network based classifier; and

the radar-based object classifier is a neural network based classifier.

19. The apparatus of claim 18 wherein the processor is configured to select between the image-based object classification and the radar-based-object classification to output as the object classification for the object automatically.