Apparatus for enhancing image quality and method for the same

1. An apparatus for enhancing image quality, the apparatus comprising:

one or more processors configured to:

combine an independent image having pixels independent from each other and a correlated image including correlation information between pixels; and

using an artificial intelligence (AI) based prediction model, provide a weight used to combine the independent image and the correlated image,

wherein the independent image and the correlated image are combined in pixel units, and

wherein the AI based prediction model is trained with the independent image and the correlated image together in order to provide the weight.

2. The apparatus of claim 1 , wherein the weight is different for each pixel.

3. The apparatus of claim 1 , wherein at least one independent image and at least one correlated image are included.

4. The apparatus of claim 1 , wherein a feature of at least one of the independent image or the correlated image is further trained.

5. The apparatus of claim 4 , wherein the feature includes at least one of normal information, texture information or depth information.

6. The apparatus of claim 5 , wherein the feature includes all of normal information, texture information and depth information.

7. The apparatus of claim 1 , wherein the weight includes:

a first weight for a pixel prediction value of a target pixel c in the independent image; and

a second weight for a sum of an independent pixel prediction value of a certain adjacent pixel i adjacent to the target pixel c and a difference z c −z i between correlated pixel prediction values of the target pixel c and the adjacent pixel i.

8. The apparatus of claim 7 , wherein the weight includes a third weight which is a divisor for a pixel prediction value, to which the first weight and the second weight are applied.

9. The apparatus of claim 8 , wherein the third weight is a sum of the first weight and the second weight.

10. The apparatus of claim 1 , wherein a prediction value of a pixel is obtained using Equation:

1

W

c

⁢

(

w

c

⁢

y

c

+

∑

i

∈

Ω

c

⁢

w

i

⁡

(

y

i

+

z

c

-

z

i

)

)

,

where, wc denotes a weight of a target pixel c, yc denotes an independent prediction value of the target pixel c, Ωc denotes a set of adjacent pixels adjacent to the target pixel, wi denotes a weight of an adjacent pixel i included in Ωc adjacent to the target pixel, W c =w c +Σ i∈Ω c w i ,y i denotes an independent prediction value of the adjacent pixel, zc denotes a correlated pixel prediction value of the target pixel, and zi denotes a correlated pixel prediction value of the adjacent pixel.

11. A method of enhancing image quality, the method comprising:

receiving at least one independent image and at least one correlated image;

obtaining, using an artificial intelligence (AI) based prediction model, a weight to be used to combine the independent image and the correlated image using the independent image and the correlated image; and

outputting a new image by combining the independent image and the correlated image according to the weight,

wherein the independent image and the correlated image are combined in pixel units, and

wherein the AI based prediction model is trained with the independent image and the correlated image together in order to provide the weight.

12. The method of claim 11 , wherein the weight includes a first weight for a pixel prediction value of a target pixel in the independent image.

13. The method of claim 12 , wherein the weight includes a second weight for a sum of an independent pixel prediction value of a certain adjacent pixel i adjacent to the target pixel c and a difference z c −z i between correlated pixel prediction values of the target pixel c and the adjacent pixels i.

14. The method of claim 13 , wherein the weight includes a third value which is a divisor for a pixel prediction value, to which the first weight and the second weight are applied.

15. The method of claim 11 , wherein the independent image, the correlated image, a feature of the independent image and a feature of the correlated image are trained, in order to obtain the weight.

16. An apparatus for enhancing image quality, the apparatus comprising:

a processor configured to:

receive an independent image having pixels independent from each other and a correlated image including correlation information between pixels;

provide a combined image by combining the independent image and the correlated image with each other; and

using an artificial intelligence (AI) based prediction model, provide a weight which is a control factor for including any one of information on the independent image or information on the correlated image in the combined image,

wherein the AI based prediction model is configured to provide the weight using both the independent image and the correlated image,

wherein the independent image and the correlated image are combined in pixel units, and

wherein the AI based prediction model is trained with the independent image and the correlated image together in order to provide the weight.

17. The apparatus of claim 16 , wherein the weight is provided in units of pixels of the image.

18. The apparatus of claim 16 ,

wherein a feature of the independent image and a feature of the correlated image are further input, and

wherein the feature is further trained in the AI based prediction model.