Image system including image signal processor and operation method of image signal processor

1. An operation method of an image signal processor (ISP) configured to perform signal processing on a raw image received from an image device, the operation method comprising:

generating a plurality of multi-scale images based on an input image, the plurality of multi-scale images having resolutions that are different from each other;

iteratively performing a fast global weighted least squares (FGWLS) based operation on each of the plurality of multi-scale images to generate a final illuminance map; and

outputting an enhanced image based on the final illuminance map and the input image.

2. The operation method of claim 1 , wherein a result of a current FGWLS based operation is transferred to an input of a next FGWLS based operation during the iteratively performing the FGWLS based operation.

3. The operation method of claim 1 , wherein the generating the plurality of multi-scale images comprises:

generating an initial illuminance map based on the input image; and

iteratively scaling down a size of the initial illuminance map n times to generate the plurality of multi-scale images, where n is a positive real number.

4. The operation method of claim 3 , wherein the iteratively performing the FGWLS based operation based on the plurality of multi-scale images to generate the final illuminance map comprises:

performing a first FGWLS based operation by using a first image having a smallest resolution from among the plurality of multi-scale images as an input image and a guidance image;

scaling up a result of the first FGWLS based operation as much as n times to generate a first scale-up result;

performing a second FGWLS based operation by using the first scale-up result as an input image, and using a second image having a resolution that is identical to a resolution of the first scale-up result, from among the plurality of multi-scale images, as a guidance image;

scaling up a result of the second FGWLS based operation as much as n times to generate a second scale-up result;

performing a third FGWLS based operation by using the second scale-up result as an input image, and using a third image having a resolution that is identical to a resolution of the second scale-up result, from among the plurality of multi-scale images, as a guidance image; and

outputting a result of the third FGWLS based operation as the final illuminance map.

5. The operation method of claim 4 , wherein the third image corresponds to the initial illuminance map.

6. The operation method of claim 1 , wherein the outputting the enhanced image based on the final illuminance map and the input image comprises:

generating a gray image based on the input image;

extracting reflectance from the gray image based on the final illuminance map;

enhancing a global contrast of the final illuminance map to generate an enhanced illuminance map;

enhancing a local contrast of the reflectance to generate enhanced reflectance; and

combining the enhanced illuminance map and the enhanced reflectance to output the enhanced image.

7. The operation method of claim 6 , wherein the enhancing the global contrast of the final illuminance map is performed based on a windowed inverse sigmoid (WIS) function.

8. The operation method of claim 6 , wherein the combining the enhanced illuminance map and the enhanced reflectance to output the enhanced image comprises:

flattening the enhanced reflectance to generate flattened reflectance;

normalizing the enhanced illuminance map to generate a normalized illuminance map; and

generating an enhanced gray image through a linear combination of the enhanced reflectance and the flattened reflectance based on the normalized illuminance map.

9. The operation method of claim 8 , wherein the flattening the enhanced reflectance comprises performing a FGWLS based operation on the enhanced reflectance.

10. The operation method of claim 8 , wherein the outputting the enhanced image further comprises:

outputting the enhanced image based on the enhanced gray image, the input image, and the gray image.

11. The operation method of claim 1 , further comprising:

receiving the raw image from the image device; and

performing pre-processing on the raw image to generate the input image.

12. The operation method of claim 11 , wherein performing the pre-processing comprises:

reducing a noise of the raw image;

performing white balancing on a result of the reducing the noise;

performing color filter array interpolation on a result of the white balancing; and

performing correction on a result of the color filter array interpolation to generate the input image.

13. An image system comprising:

an image sensor configured to output a raw image based on a light reflected from an external object; and

an image signal processor configured to:

perform pre-processing on the raw image to generate an input image,

extract a final illuminance map from the input image through a multi-scale based extraction operation, and

generate an enhanced image based on the final illuminance map and the input image,

wherein the multi-scale based extraction operation comprises:

generating a plurality of multi-scale images having different resolutions based on the input image;

iteratively performing an illuminance map extraction operation on the plurality of multi-scale images; and

outputting the final illuminance map as a result of iteratively performing the illuminance map extraction operation.

14. The image system of claim 13 , wherein the plurality of multi-scale images are generated by iteratively scaling down an initial illuminance map, which is generated based on the input image, as much as n times to have the different resolutions, where n is a positive integer greater than 1.

15. The image system of claim 14 , wherein the illuminance map extraction operation is performed based on:

L

i

(

x

,

y

)

=

{

FGWLS

(

↑

n

(

L

i

+

1

(

x

,

y

)

)

,

G

i

(

x

,

y

)

,

λ

i

,

T

i

)

,

i

=

0

,

…

,

l

-

1

FGWLS

⁢

(

L

ˆ

i

(

x

,

y

)

,

G

i

(

x

,

y

)

,

λ

i

,

T

i

)

,

i

=

l

where (x,y) is a pixel location, FGWLS is a function for a fast global weighted least squares (FGWLS) based operation, L i is a calculation result of the FGWLS, ↑ n is a function to scale up an input as much as the n times, {circumflex over (L)} i is an illuminance map obtained by scaling down the initial illuminance map as much as n l times, G i is a guidance image, λ i is a flattening level used in the function of the FGWLS based operation, and T i is a number of times of row-based calculation and column-based calculation used in the function for the FGWLS based operation, and

wherein the illuminance map extraction operation is performed based on the i that stepwise changes from 1 to 0.

16. The image system of claim 15 , wherein the guidance image is a corresponding image of the plurality of multi-scale images.

17. An image contrast enhancing method of an image signal processor, the image contrast enhancing method comprising:

generating an initial illuminance map based on an input image;

iteratively scaling down the initial illuminance map as much as n times to generate a plurality of multi-scale images, where n is a positive integer greater than 1;

performing a first calculation based on a first image having a lowest resolution from among the plurality of multi-scale images;

performing a second calculation based on a result of the first calculation and a second image having a second lowest resolution from among the plurality of multi-scale images;

performing a third calculation based on a result of the second calculation and a third image having a highest resolution from among the plurality of multi-scale images;

generating a result of the third calculation as a final illuminance map; and

generating an enhanced image based on the final illuminance map and the input image,

wherein the first calculation, the second calculation, and the third calculation are based on fast global weighted least squares (FGWLS).

18. The image contrast enhancing method of claim 17 , wherein the first calculation is performed based on a first input, a first guidance image, a first flattening level, and a first iteration count,

wherein the second calculation is performed based on a second input, a second guidance image, a second flattening level, and a second iteration count,

wherein the third calculation is performed based on a third input, a third guidance image, a third flattening level, and a third iteration count,

wherein the first input is the first image,

wherein the second input is an image obtained by enlarging the result of the first calculation as much as the n times,

wherein the third input is an image obtained by enlarging the result of the second calculation as much as the n times,

wherein the first guidance image is the first image,

wherein the second guidance image is the second image, and

wherein the third guidance image is the third image.

19. The image contrast enhancing method of claim 17 , wherein the generating the enhanced image based on the final illuminance map and the input image comprises:

extracting reflectance from a gray image corresponding to the input image based on the final illuminance map;

enhancing a global contrast of the final illuminance map to generate an enhanced illuminance map;

enhancing a local contrast of the reflectance to generate enhanced reflectance; and

combining the enhanced illuminance map and the enhanced reflectance to output an enhanced gray image.

20. The image contrast enhancing method of claim 19 , wherein the generating the enhanced image based on the final illuminance map and the input image further comprises:

recovering a color of the enhanced gray image based on the input image to generate the enhanced image.