Method for camera control, image signal processor and device

1. A method for camera control to acquire an image, comprising:

acquiring a stream of image frames by an image sensor comprising at least one frame;

acquiring a target frame by the image sensor;

determining scene information of the target frame;

selecting at least one reference frame from the stream of image frames by identifying the scene information of the target frame in the reference frame;

determining at least one acquisition parameter of the reference frame; and

determining a final image from the target frame by correcting at least one corresponding acquisition parameter of the target frame using the at least one acquisition parameter of the reference frame,

wherein the scene information comprises simultaneous localization and mapping (SLAM) data for the image frame of the stream of image frames and the target frame.

2. The method according to claim 1 , wherein acquisition parameter is related to at least one of an auto white balancing (AWB), automatic exposure control (AEC) or tone-mapping (TM) parameter.

3. The method according to claim 1 , wherein the scene information comprises depth information of the image frame and/or the target frame and/or pose of the image sensor.

4. The method according to claim 1 , wherein selecting a reference frame from the stream of image frames by identifying the scene information of the target frame in the reference frame comprises determining an at least partial overlap of the image frame from the stream of image frames with the target frame by the scene information.

5. The method according to claim 4 , wherein scene information comprise coordinates of the scene, and wherein selecting a reference frame from the stream of images by identifying the scene information of the target frame comprises calculating coordinates of the target frame and determining an at least partial overlap with coordinates in the respective image frames of the stream of image frames.

6. The method according to claim 5 , wherein calculating coordinates of the scene comprises:

acquiring depth information d for pixel in the frame (u, v, 0);

determining coordinates in the camera system (X cam , Y cam , d, 1) by

X cam =( u× 4− px )× d÷cx , and

Y cam =( v× 4+60− py )× d÷cy,

with (px, py) being the principal point of the image sensor, (cx, cy) being the focal length; and

transferring the coordinates to the world coordinate system by

(

X

Y

Z

1

)

=

(

R

⁢

❘

"\[LeftBracketingBar]"

t

)

·

(

X

cam

Y

cam

d

1

)

,

with (X, Y, Z, 1) being the coordinates in the world coordinate system and (R|t) the pose of the image sensor.

7. The method according to claim 6 , comprising comparing the coordinates in the world coordinate system of the object in the target frame with each image frame from the stream to determine the at least partial overlap.

8. The method according to claim 1 , wherein selecting a reference frame comprises determining a confidence level of the respective frame for the acquisition parameter and selecting the reference frame if the confidence level is above a preset threshold.

9. The method according to claim 1 , wherein the reference frame is selected by the maximum of overlap and confidence value of the respective image frame of the stream of image frames.

10. The method according to claim 9 , wherein the confidence value is provided by one or more of color gamut, brightness gamut, a hull of the 2D chromaticity, 1D brightness range, or 3D color histogram.

11. The method according to claim 1 , wherein the image frames from the stream of image frames comprises low resolution images having a resolution lower than the final image or 3A statistics of the raw image frame.

12. The method according to claim 1 , wherein the image frames of the stream of image frames are stored.

13. The method according to claim 1 , wherein image frames of the stream of image frames are stored having a confidence level above a preset threshold.

14. The method according to claim 12 , wherein camera pose is stored together with the stored image frames of the stream of image frames.

15. The method according to claim 1 , wherein the method further comprises:

detecting change of illumination between the reference frame and the target frame and adapting the reference frame to the changed illumination before determining the acquisition parameter.

16. The method according to claim 1 , wherein steps of the method are repeated for every new target frame of a video stream or a stream of preview-images.

17. An image signal processor (ISP), configured to:

acquire a stream of image frames by an image sensor comprising at least one frame;

acquire a target frame by the image sensor;

determine scene information of the target frame;

select at least one reference frame from the stream of image frames by identifying the scene information of the target frame in the reference frame;

determine at least one acquisition parameter of the reference frame; and

determine a final image from the target frame by correcting at least one corresponding acquisition parameter of the target frame using the at least one acquisition parameter of the reference frame,

wherein the scene information comprises simultaneous localization and mapping (SLAM) data for the image frame of the stream of image frames and the target frame.

18. The ISP according to claim 17 , wherein acquisition parameter is related to at least one of an auto white balancing (AWB), automatic exposure control (AEC) or tone-mapping (TM) parameter.

19. A camera device comprising: a processor and a non-transitory memory storage storing instruction, when executed by the processor, causing the processor to perform acts comprising:

acquiring a stream of image frames by an image sensor comprising at least one frame;

acquiring a target frame by the image sensor;

determining scene information of the target frame;

selecting at least one reference frame from the stream of image frames by identifying the scene information of the target frame in the reference frame;

determining at least one acquisition parameter of the reference frame; and

determining a final image from the target frame by correcting at least one corresponding acquisition parameter of the target frame using the at least one acquisition parameter of the reference frame,

wherein the scene information comprises simultaneous localization and mapping (SLAM) data for the image frame of the stream of image frames and the target frame.

20. The camera device according to claim 19 , wherein acquisition parameter is related to at least one of an auto white balancing (AWB), automatic exposure control (AEC) or tone-mapping (TM) parameter.