Pixel consistency

1. A method of image processing, the method comprising:

receiving image data;

applying image processing to the image data to produce a processed image;

producing consistency data associated with the image processing, wherein the consistency data is indicative of a consistency of at least one region of the processed image with the received image data; and

using both the processed image and the consistency data as inputs to a decision process of a computer vision system.

2. The method according to claim 1 , comprising receiving the image data from an image sensor, the image data being captured by the image sensor.

3. The method according to claim 1 , wherein the image data comprises a video frame.

4. The method according to claim 1 , wherein:

the image processing comprises correction of at least one defective pixel in the image data; and

the consistency data identifies at least one corrected pixel in the processed image corresponding to the at least one defective pixel in the image data.

5. The method according to claim 1 , wherein:

the processed image is a high dynamic range image:

the image processing comprises producing the high dynamic range image from a plurality of images at varying exposure settings; and

the consistency data identifies at least one region in the processed image corresponding to at least one region in which there is an inconsistency between the plurality of images.

6. The method according to claim 1 , wherein:

the image processing comprises identifying and correcting for a flickering light source in the image data; and

the consistency data identifies at least one corrected region in the processed image corresponding to the flickering light source in the image data.

7. The method according to claim 1 , comprising navigating an autonomous vehicle based on the decision process of the computer vision system.

8. The method according to claim 1 , wherein a region of the at least one region is a single pixel of the processed image.

9. The method according to claim 1 , wherein a region of the at least one region is a group of pixels of the processed image, the group of pixels having predetermined dimensions.

10. The method according to claim 9 , wherein the predetermined dimensions depend on a position of the group of pixels within the processed image.

11. The method according to claim 1 , comprising storing the consistency data as a layer of the processed image.

12. A non-transitory computer-readable storage medium comprising a set of computer-readable instructions stored thereon which, when executed by at least one processor, cause the at least one processor to:

receive image data;

apply image processing to the image data to produce a processed image;

produce consistency data associated with the image processing, wherein the consistency data is indicative of a consistency of at least one region of the processed image with the received image data; and

provide both the processed image and the consistency data as inputs to a decision process of a computer vision system.

13. A computer vision apparatus comprising at least one processor, the at least one processor being configured to:

receive a processed image and consistency data associated with image processing applied to image data whereby to produce the processed image, the consistency data indicating a consistency of at least one region of the processed image with the received image data; and

based on the processed image and the consistency data, perform a computer vision decision process.

14. The computer vision apparatus according to claim 13 , further comprising an image sensor configured to capture the image data.

15. The computer vision apparatus according to claim 13 , wherein the at least one processor is configured to:

receive the image data;

apply image processing to the image data, whereby to produce the processed image; and

generate the consistency data.

16. The computer vision apparatus according to claim 13 , wherein:

the computer vision apparatus is associated with an autonomous vehicle; and

the decision process is an obstacle avoidance process of the autonomous vehicle.

17. The computer vision apparatus according to claim 16 , wherein the decision process comprises decreasing a weight of input data derived from the processed image and increasing a weight of input data derived from a source other than the processed image.