Methods for analysis of an image and a method for generating a dataset of images for training a machine-learned model
A method for analysis of an image comprises: receiving ( 402 ) the image to be analyzed; processing ( 404 ) the image with a machine-learned model, wherein the machine-learned model is configured to predict at least an intrinsic parameter of the image using at least a first variable of the machine-learned model, wherein the first variable defines a relation between a radial distortion of the image and a focal length of the image; and outputting ( 406 ) the intrinsic parameter of the image. Also, methods for forming a 3D reconstruction of a scenery, for training a machine-learned model for analysis of an image and for generating a dataset of images for training a machine-learned model are disclosed.
1. A method for analysis of an image, said method comprising:
receiving the image to be analyzed;
processing the image with a machine-learned model, wherein the machine-learned model is configured to predict at least an intrinsic parameter of the image using at least a first variable of the machine-learned model, wherein the first variable defines a relation between a radial distortion of the image and a focal length of the image; and
outputting the intrinsic parameter of the image.
2. The method according to claim 1 , wherein the relation defined by the first variable is proportional to the radial distortion and inversely proportional to a square of the focal length.
3. The method according to claim 1 , wherein the first variable is defined as:
k
1
f
2
wherein is the first variable, k 1 is a radial distortion parameter of the image and f is the focal length of the image.
4. The method of claim 1 , wherein the machine-learned model is configured to use a pre-set relation between a first radial distortion parameter and a second radial distortion parameter.
5. The method of claim 4 , wherein the machine-learned model is further configured to predict an uncertainty of the intrinsic parameter and the method further comprises outputting the uncertainty.
6. The method of claim 1 , wherein the machine-learned model being configured to predict at least an intrinsic parameter of the image comprises the machine-learned model being configured to predict a field of view of the image.
7. The method of claim 6 , wherein the machine-learned model is further configured to predict at least an extrinsic parameter of a camera using which the received image has been acquired.
8. The method of claim 1 , further comprising correcting coordinates of feature points in the image based on the outputted intrinsic parameter.
9. A computer program product, the computer program product being embodied in a non-transitory computer readable storage medium and comprising computer-readable instructions for:
receiving the image to be analyzed;
processing the image with a machine-learned model, wherein the machine-learned model is configured to predict at least an intrinsic parameter of the image using at least a first variable of the machine-learned model, wherein the first variable defines a relation between a radial distortion of the image and a focal length of the image; and
outputting the intrinsic parameter of the image.
10. The computer program product of claim 9 , wherein the relation defined by the first variable is proportional to the radial distortion and inversely proportional to a square of the focal length.
11. The computer program product of claim 9 , wherein the machine-learned model is configured to use a pre-set relation between a first radial distortion parameter and a second radial distortion parameter.
12. The computer program product of claim 9 , wherein the machine-learned model being configured to predict at least an intrinsic parameter of the image comprises the machine-learned model being configured to predict a field of view of the image.
13. A system for analysis of an image, comprising:
a processor configured to:
receive the image to be analyzed;
process the image with a machine-learned model, wherein the machine-learned model is configured to predict at least an intrinsic parameter of the image using at least a first variable of the machine-learned model, wherein the first variable defines a relation between a radial distortion of the image and a focal length of the image; and
output the intrinsic parameter of the image; and
a memory coupled to the processor and configured to provide the processor with instructions.
14. The system of claim 13 , wherein the relation defined by the first variable is proportional to the radial distortion and inversely proportional to a square of the focal length.
15. The system of claim 13 , wherein the first variable is defined as:
=
k
1
f
2
wherein is the first variable, k 1 is a radial distortion parameter of the image and f is the focal length of the image.
16. The system of claim 13 , wherein the machine-learned model is configured to use a pre-set relation between a first radial distortion parameter and a second radial distortion parameter.
17. The system of claim 16 , wherein the machine-learned model is further configured to predict an uncertainty of the intrinsic parameter and the method further comprises outputting the uncertainty.
18. The system of claim 13 , wherein the machine-learned model being configured to predict at least an intrinsic parameter of the image comprises the machine-learned model being configured to predict a field of view of the image.
19. The system of claim 18 , wherein the machine-learned model is further configured to predict at least an extrinsic parameter of a camera using which the received image has been acquired.
20. The system of claim 13 , wherein the processor is further configured to correct coordinates of feature points in the image based on the outputted intrinsic parameter.