Asynchronous object ROI detection in video mode

1. An image data processing method comprising:

receiving frame image data of a plurality of frames;

receiving a plurality of control rules comprising a respective control rule for each of the frames, wherein each of the control rules identifies one of a plurality of process regions and one of a plurality of object detectors;

identifying a region of interest in each frame by a location and a category, comprising applying the object detector identified by the respective control rule to the respective frame image data in the process region identified by the respective control rule;

identifying a final region of interest based on the identified regions of interest; and

reporting the final region of interest.

2. The image data processing method of claim 1 , wherein identifying a final region of interest comprises:

identifying a final location of the final region of interest based on the locations of the regions of interest.

3. The image data processing method of claim 1 , wherein identifying a region of interest in each frame comprises:

identifying a category confidence value, wherein the category confidence value represents a confidence in the identified category.

4. The image data processing method of claim 1 , wherein identifying a final region of interest comprises:

identifying a category of the final region of interest based on the categories of the regions of interest.

5. The image data processing method of claim 4 , wherein identifying a category of the final region of interest comprises: selecting one of the categories according to category confidence values.

6. The image data processing method of claim 5 , wherein selecting one of the categories comprises:

selecting the one of the categories having the highest category confidence value.

7. An image sensor comprising:

a frame input unit configured to receive, from the image sensor, frame image data of a plurality of frames;

a rule input unit configured to receive a plurality of control rules comprising a respective control rule for each of the frames, wherein each of the control rules identifies one of a plurality of process regions and one of a plurality of object detectors;

a process unit configured to identify a region of interest in each frame by a location and a category, the process unit further configured to comprising applying the object detector identified by the respective control rule to the respective frame image data in the process region identified by the respective control rule;

a vote unit configured to identify a final region of interest based on the identified regions of interest; and

a report unit configured to report the final region of interest.

8. The image sensor of claim 7 , wherein the vote unit is further configured to:

identify a final location of the final region of interest based on the locations of the regions of interest.

9. The image sensor of claim 7 , wherein the process unit is further configured to:

identify a category confidence value, wherein the category confidence value represents a confidence in the identified category.

10. The image sensor of claim 7 , wherein the vote unit is further configured to:

identify a category of the final region of interest based on the categories of the regions of interest.

11. The image sensor of claim 10 ,

wherein to identify the category of the final region of interest, the vote unit is further configured to: select one of the categories according to category confidence values.

12. The image sensor of claim 11 , wherein to select one of the categories, the vote unit is further configured to:

select the one of the categories having the highest category confidence value.

13. A non-transitory machine-readable storage medium encoded with instructions executable by a hardware processor, the machine-readable storage medium comprising instructions to cause the hardware processor to perform an image data processing method, the method comprising:

receiving, from an image sensor, frame image data of a plurality of frames;

receiving a plurality of control rules comprising a respective control rule for each of the frames, wherein each of the control rules identifies one of a plurality of process regions and one of a plurality of object detectors;

identifying a region of interest in each frame by a location and a category, comprising applying the object detector identified by the respective control rule to the respective frame image data in the process region identified by the respective control rule;

identifying a final region of interest based on the identified regions of interest; and

reporting the final region of interest.

14. The medium of claim 13 , wherein identifying a final region of interest comprises:

identifying a final location of the final region of interest based on the locations of the regions of interest.

15. The medium of claim 13 , wherein identifying a region of interest in each frame comprises:

identifying a category confidence value, wherein the category confidence value represents a confidence in the identified category.

16. The medium of claim 13 , wherein identifying a final region of interest comprises:

identifying a category of the final region of interest based on the categories of the regions of interest.

17. The medium of claim 16 ,

wherein identifying a category of the final region of interest comprises: selecting one of the categories according to category confidence values.

18. The medium of claim 17 , wherein selecting one of the categories comprises:

selecting the one of the categories having the highest category confidence value.