Image processing method and system

1. A computer-implemented method for image processing, comprising:

determining a rotated angle of an identification (ID) document image, wherein the rotated angle corresponds to an angle between a bottom edge of the ID document image and an x-axis;

rotating the ID document image based on the rotated angle, wherein an angle is defined between the bottom edge of the rotated ID document image and the x-axis, and wherein a value of the angle is within a predefined range;

performing image regression on the rotated ID document image to obtain coordinates of four corner points of the rotated ID document image; and

performing image alignment on the rotated ID document image based on the coordinates of four corner points of the rotated ID document image.

2. The computer-implemented method of claim 1 , wherein the rotated angle of the ID document image is determined using a trained angle-determining model.

3. The computer-implemented method of claim 2 , wherein the trained angle-determining model is trained using an angle-determining regression task of an angle-determining convolutional neural network.

4. The computer-implemented method of claim 1 , wherein the predefined range of the angle between the bottom edge of the rotated ID document image and the x-axis is between −10 degrees and 10 degrees.

5. The computer-implemented method of claim 1 , wherein the image regression on the rotated ID document image is performed using a trained document detection model.

6. The computer-implemented method of claim 5 , wherein the trained document detection model is trained using a document detection regression task of a document detection convolutional neural network.

7. The computer-implemented method of claim 1 , wherein performing image alignment on the rotated ID document image comprises performing an image warping task.

8. A non-transitory computer-readable medium storing one or more instructions executable by a computer system to perform operations for image processing, comprising:

determining a rotated angle of an identification (ID) document image, wherein the rotated angle corresponds to an angle between a bottom edge of the ID document image and an x-axis;

rotating the ID document image based on the rotated angle, wherein an angle is defined between the bottom edge of the rotated ID document image and the x-axis, and wherein a value of the angle is within a predefined range;

performing image regression on the rotated ID document image to obtain coordinates of four corner points of the rotated ID document image; and

performing image alignment on the rotated ID document image based on the coordinates of four corner points of the rotated ID document image.

9. The non-transitory computer-readable medium of claim 8 , wherein the rotated angle of the ID document image is determined using a trained angle-determining model.

10. The non-transitory computer-readable medium of claim 9 , wherein the trained angle-determining model is trained using an angle-determining regression task of an angle-determining convolutional neural network.

11. The non-transitory computer-readable medium of claim 8 , wherein the predefined range of the angle between the bottom edge of the rotated ID document image and the x-axis is between −10 degrees and 10 degrees.

12. The non-transitory computer-readable medium of claim 8 , wherein the image regression on the rotated ID document image is performed using a trained document detection model.

13. The non-transitory computer-readable medium of claim 12 , wherein the trained document detection model is trained using a document detection regression task of a document detection convolutional neural network.

14. The non-transitory computer-readable medium of claim 8 , wherein performing image alignment on the rotated ID document image comprises performing an image warping task.

15. A computer-implemented system for image processing, comprising:

one or more computers; and

one or more computer memory devices interoperably coupled with the one or more computers and having tangible, non-transitory, machine-readable media storing one or more instructions that, when executed by the one or more computers, cause the one or more computers to perform one or more operations comprising:

determining a rotated angle of an identification (ID) document image, wherein the rotated angle corresponds to an angle between a bottom edge of the ID document image and an x-axis;

rotating the ID document image based on the rotated angle, wherein an angle is defined between the bottom edge of the rotated ID document image and the x-axis, and wherein a value of the angle is within a predefined range;

performing image regression on the rotated ID document image to obtain coordinates of four corner points of the rotated ID document image; and

performing image alignment on the rotated ID document image based on the coordinates of four corner points of the rotated ID document image.

16. The computer-implemented system of claim 15 , wherein the rotated angle of the ID document image is determined using a trained angle-determining model.

17. The computer-implemented system of claim 16 , wherein the trained angle-determining model is trained using an angle-determining regression task of an angle-determining convolutional neural network.

18. The computer-implemented system of claim 15 , wherein the predefined range of the angle between the bottom edge of the rotated ID document image and the x-axis is between −10 degrees and 10 degrees.

19. The computer-implemented system of claim 15 , wherein the image regression on the rotated ID document image is performed using a trained document detection model.

20. The computer-implemented system of claim 19 , wherein the trained document detection model is trained using a document detection regression task of a document detection convolutional neural network.