Electronic apparatus and method for controlling thereof

1. A method for controlling an electronic apparatus comprising:

inputting a first light field (LF) image set, the first LF image set comprising a plurality of view images of different viewpoints of a same object, to a first artificial intelligence model to acquire a pixel shift value including a sub-pixel shift value corresponding to each pixel for converting pixels in the first LF image set;

acquiring a second LF image set by converting the pixels in the first LF image set according to the pixel shift value;

inputting the first LF image set and the second LF image set to a second artificial intelligence model for converting the LF image sets to a layer image to acquire the layer image;

inputting the acquired layer image to a simulation model for restoring the LF image sets to acquire a third LF image set; and

transmitting, to an external server, learning data for learning the first artificial intelligence model and the second artificial intelligence model based on the second LF image set and the third LF image set.

2. The method of claim 1 , wherein the acquiring of the second LF image set comprises acquiring the second LF image set by moving, according to a pixel shift value corresponding to each of the plurality of view images included in the first LF image set, pixels of each of the plurality of view images.

3. The method of claim 2 , wherein the acquiring of the second LF image set comprises acquiring a first view image of the second LF image set by moving, according to a first pixel shift value corresponding to a first view image, pixels in the first view image included in the first LF image set.

4. The method of claim 3 ,

wherein the acquiring of the first view image of the second LF image set comprises acquiring the first view image of the second LF image set by moving each pixel in the first view image included in the first LF image set according to the sub-pixel shift value included in the first pixel shift value.

5. The method of claim 1 , wherein the acquiring of the layer image comprises inputting the first LF image set, the second LF image set, and the pixel shift value to the second artificial intelligence model to acquire the layer image.

6. The method of claim 1 ,

wherein the layer image is configured to include a first layer image, a second layer image, and a third layer image, and

wherein the acquiring of the third LF image set comprises:

shifting each of the first layer image and the third layer image for each view point, and

acquiring the third LF image set by cropping the shifted first layer image and the shifted third layer image together with the second layer image.

7. The method of claim 1 , wherein the learning comprises:

acquiring a loss function by comparing the second LF image set with the third LF image set; and

learning the first artificial intelligence model and the second artificial intelligence model based on the loss function.

8. The method of claim 1 , further comprising:

inputting the first LF image set to the learned first artificial intelligence model to acquire the pixel shift value;

acquiring the second LF image set by converting the pixels in the first LF image set according to the pixel shift value; and

inputting the first LF image set and the second LF image set to the learned second artificial intelligence model to acquire a layer image.

9. The method of claim 1 , further comprising:

inputting the first LF image set to the learned first artificial intelligence model to acquire the pixel shift value; and

inputting the first LF image set and the pixel shift value to the learned second artificial intelligence model to acquire the layer image.

10. The method of claim 8 , further comprising:

providing the layer image to a stacked display.

11. An electronic apparatus comprising:

a memory for storing at least one instruction; and

a processor for executing the at least one instruction stored in the memory to control the electronic apparatus,

wherein the processor is configured to:

input a first light field (LF) image set, the first LF image set comprising a plurality of view images of different viewpoints of a same object, to a first artificial intelligence model to acquire a pixel shift value including a sub-pixel shift value corresponding to each pixel for converting pixels in the first LF image set,

acquire a second LF image set by converting the pixels in the first LF image set according to the pixel shift value,

input the first LF image set and the second LF image set to a second artificial intelligence model for converting the LF image set to a layer image to acquire the layer image,

input the acquired layer image to a simulation model for restoring the LF image sets to acquire a third LF image set, and

transmit, to an external server, learning data to learn the first artificial intelligence model and the second artificial intelligence model based on the second LF image set and the third LF image set.

12. The electronic apparatus of claim 11 , wherein the processor is further configured to acquire the second LF image set by moving, according to a pixel shift value corresponding to each of the plurality of view images included in the first LF image set, pixels of each of the plurality of view images.

13. The electronic apparatus of claim 12 , wherein the processor is further configured to acquire a first view image of the second LF image set by moving, according to a first pixel shift value corresponding to a first view image, pixels in the first view image included in the first LF image set.

14. The electronic apparatus of claim 13 ,

wherein the processor is further configured to acquire the first view image of the second LF image set by moving each pixel in the first view image included in the first LF image set according to the sub-pixel shift value included in the first pixel shift value.

15. The electronic apparatus of claim 11 , wherein the processor is further configured to input the first LF image set, the second LF image set, and the pixel shift value to the second artificial intelligence model to acquire the layer image.

16. The electronic apparatus of claim 11 , wherein the restoring of the LF image set to acquire a third LF image set comprises setting an average value of pixel values of an overlapping area of a plurality of layer images to a pixel value of a corresponding area.

17. The electronic apparatus of claim 11 , wherein the restoring of the LF image set to acquire a third LF image set comprises setting a value obtained by multiplying pixel values of overlapping areas of a plurality of layer images as a pixel value of a corresponding area.

18. The electronic apparatus of claim 11 , wherein the restoring of the LF image set to acquire a third LF image set comprises setting a value obtained by adding pixel values of overlapping areas of a plurality of layer images as a pixel value of a corresponding area.