Method and device for recovering image passing through display
View Patent ↗A method for recovering an image passing through a display and taken by an image sensor disposed on a rear side of the display, the method includes acquiring the image taken by the image sensor, obtaining a plurality of point spread functions corresponding one-to-one to a plurality of dots arranged in a predetermined pattern on a calibration chart, and recovering the image by performing a de-convolution of the image based on the plurality of point spread functions.
1. A method for recovering an image passing through a display and taken by an image sensor disposed on a rear side of the display, the method comprising:
acquiring the image taken by the image sensor;
obtaining a plurality of point spread functions corresponding one-to-one to a plurality of dots arranged in a predetermined pattern on a calibration chart; and
recovering the image by performing a de-convolution of the image based on the plurality of point spread functions.
2. The method according to claim 1 , wherein the recovering the image by performing the de-convolution of the image based on the plurality of point spread functions to comprises:
obtaining, for each pixel of the image sensor, a PSF matrix corresponding to a pixel of the image sensor based on the plurality of point spread functions;
calculating an inverse matrix for each of the obtained PSF matrices; and
convolving the image taken by the image sensor and the inverse matrices.
3. The method according to claim 2 , wherein the obtaining, for each pixel of the image sensor, the PSF matrix corresponding to the pixel of the image sensor based on the plurality of point spread functions comprising:
selecting, for each pixel of the image sensor, a PSF matrix indicating a point spread function nearest to the pixel among the plurality of point spread functions.
4. The method according to claim 2 , wherein the obtaining, for each pixel of the image sensor, the PSF matrix corresponding to a pixel of the image sensor based on the plurality of point spread functions comprising:
creating, for each pixel of the image sensor, a PSF matrix by interpolating at least two point spread functions in a vicinity of the pixel among the plurality of point spread functions.
5. The method according to claim 1 , wherein the display has a wiring structure formed in a lattice shape, and
a distance between the display and the image sensor, a size of a substantially rectangular opening of the wiring structure, and a wavelength of light passing through the display meet a condition of Fresnel diffraction.
6. The method according to claim 5 , wherein the distance is between 0.5 mm and 7.5 mm, a length of one side of the opening is between 10 um and 200 um, and the wavelength is a wavelength of visible light.
7. The method according to claim 1 , wherein the plurality of point spread functions are obtained for each of a plurality of color channels.
8. The method according to claim 7 , wherein the plurality of color channels include Red channel, Green channel and Blue channel.
9. The method according to claim 1 , wherein in response to the recovering the image by performing a de-convolution of the image based on the plurality of point spread functions, the method further comprises:
reducing noise of the recovered image by calculating spatial filtering.
10. The method according to claim 1 , further comprising:
prior to the recovering the image by performing the de-convolution of the image based on the plurality of point spread functions, increasing digital gain of recovered image based on a light transmission rate of the display.
11. The method according to claim 1 , further comprising:
prior to the obtaining the plurality of point spread functions corresponding one-to-one to the plurality of dots arranged in the predetermined pattern on the calibration chart:
obtaining a distance between the display and the image sensor;
obtaining a grating size of a wiring structure; and
calculating a plurality of first point spread functions for each color channel based on model data of the calibration chart, the distance and the grating size.
12. The method according to claim 11 , further comprising:
obtaining a plurality of second point spread functions for each color channel by taking an image of the calibration chart by the image sensor in the entire field of view (FOV) of the image sensor; and
obtaining a plurality of third point spread functions for each color channel based on the plurality of first point spread functions and the plurality of second point spread functions.
13. The method according to claim 12 , wherein the third point spread function is obtained by taking a weighted or simple average of the first point spread function and the second point spread function in the same grid.
14. The method according to claim 12 , wherein the third point spread function is obtained by taking the larger value of the first point spread function and the second point spread function in the same grid.
15. An electronic device comprising:
a display;
an image sensor disposed on a rear side of the display;
a processor; and
a memory, wherein the processor is configured to recover an image which passes through the display and taken by the image sensor by performing acts of:
acquiring the image taken by the image sensor;
obtaining a plurality of point spread functions corresponding one-to-one to a plurality of dots arranged in a predetermined pattern on a calibration chart; and
recovering the image by performing a de-convolution of the image based on the plurality of point spread functions.
16. The electronic device according to claim 15 , wherein the processor is configured to perform a de-convolution of the image based on the plurality of point spread functions to recover the image by performing acts of:
obtaining, for each pixel of the image sensor, a PSF matrix corresponding to a pixel of the image sensor based on the plurality of point spread functions;
calculating an inverse matrix for each of the obtained PSF matrices; and
convolving the image taken by the image sensor and the inverse matrices.
17. The electronic device according to claim 15 , wherein the display has a wiring structure formed in a lattice shape, and
a distance between the display and the image sensor, a size of a substantially rectangular opening of the wiring structure, and a wavelength of light passing through the display meet a condition of Fresnel diffraction.
18. The electronic device according to claim 17 , wherein the distance is between 0.5 mm and 7.5 mm, a length of one side of the opening is between 10 um and 200 um, and the wavelength is a wavelength of visible light.
19. A non-transitory computer-readable storage medium, having instructions stored thereon that, when executed by a processor, cause the processor to implement a method for recovering an image passing through a display and taken by an image sensor disposed on a rear side of the display, the method comprising:
acquiring the image taken by the image sensor;
obtaining a plurality of point spread functions corresponding one-to-one to a plurality of dots arranged in a predetermined pattern on a calibration chart; and
recovering the image by performing a de-convolution of the image based on the plurality of point spread functions.
20. The non-transitory computer-readable storage medium according to claim 19 , wherein the recovering the image by performing the de-convolution of the image based on the plurality of point spread functions to comprises:
obtaining, for each pixel of the image sensor, a PSF matrix corresponding to a pixel of the image sensor based on the plurality of point spread functions;
calculating an inverse matrix for each of the obtained PSF matrices; and
convolving the image taken by the image sensor and the inverse matrices.