Apparatus and method for creating and providing mosaic image based on image tag-word

1. A method for creating a mosaic image based on an image tag-word by a server for providing a mosaic service, the method comprising:

determining each tag-word of each image, creating one or more categories, each category corresponding to at least one tag-word, and classifying each of a plurality of images into one of the categories according to the at least one tag-word corresponding to each category and the determined each tag-word of each image;

determining a target image among the plurality of images;

selecting a pixel image for mosaicizing the determined target image based on tag-words of the plurality of images; and

creating a mosaic image for the target image based on the selected pixel image,

wherein the determining of the target image among the plurality of images comprises:

comparing representative factor distribution data of one or more images classified for each category with representative factor distribution data of unit cells of each of the one or more images classified for each category,

selecting in each category an image according to a matching rate of the representative factor distribution data of the one or more images classified for each category and the representative factor distribution data of the unit cells of each of the one or more images classified for each category, and

providing the selected image as a recommended target image.

2. The method of claim 1 , wherein the determining of each tag-word of each image and the classifying of the plurality of images according to the determined each tag-word of each image includes inputting the plurality of images into a deep learning neural network to automatically determine each tag-word of each image.

3. The method of claim 1 ,

the recommended target image for each category among the one or more images classified for each category is selected using a deep learning neural network, and

the determining of the target image among the plurality of images further comprises acquiring an input of a user of selecting the target image from the recommended target image by providing a user interface.

4. The method of claim 1 , wherein the selecting in each category the image according to the matching rate comprises;

selecting in each category an image, in which the matching rate of the representative factor distribution data of the one or more images classified for each category and the representative factor distribution data of the unit cells of each of the one or more images classified for each category is equal to or higher than a predetermined criterion.

5. The method of claim 1 , wherein the selecting the pixel image includes:

providing an interface configured to receive selection of a tag-word or category from a user, and

selecting one or more images which correspond to the selection of the tag-word or category as the pixel image.

6. The method of claim 5 , wherein the selecting the pixel image further includes providing a recommended tag-word using a deep learning neural network through the interface.

7. The method of claim 1 , wherein the creating of the mosaic image includes dividing the target image into the unit cells, matching the pixel image for each unit cell, and inserting the pixel image matched with the each unit cell.

8. A method for creating a mosaic image based on an image tag-word by a server for providing a mosaic service, the method comprising:

determining each tag-word of each image and classifying a plurality of images according to the determined each tag-word of each image;

determining a pixel image among the plurality of images;

selecting a target image for mosaicizing the determined pixel image based on tag-words of the plurality of images; and

creating a mosaic image based on the determined pixel image and the target image,

wherein the selecting of the target image comprises:

comparing representative factor distribution data of the determined pixel image with representative factor distribution data of unit cells of the determined pixel image, and

providing the pixel image as a recommended pixel image according to a matching rate of the representative factor distribution data of the determined pixel image and the representative factor distribution data of the unit cells of the determined pixel image is equal to or higher than a predetermined criterion.

9. The method of claim 8 , wherein:

the recommended target image among the determined pixel image and/or the plurality of images is selected using a deep learning neural network, and

the selecting of the target image comprises acquiring an input of a user of selecting the target image based on the recommended target image and/or the plurality of images.

10. The method of claim 8 , wherein:

the providing of the pixel image as the recommended pixel image according to the matching rate comprises providing the pixel image, in which a matching rate of the representative factor distribution data of the determined pixel image and the representative factor distribution data of the unit cells of the determined pixel image is equal to or higher than a predetermined criterion, as the recommended pixel image.

11. An apparatus for providing a mosaic image, the apparatus comprising:

a storage configured to store a plurality of images and a program for creating a mosaic image based on the plurality of images;

a display configured to output the created mosaic image; and

a processor configured to read the program stored in the storage and create the mosaic image,

wherein the processor is configured to:

determine each tag-word of each image, create one or more categories, each category corresponding to at least one tag-word, and classify each of the plurality of images into one of the categories according to the at least one tag-word corresponding to each category and the determined each tag-word of each image,

determine a target image among the plurality of images by comparing representative factor distribution data of one or more images classified for each category with representative factor distribution data of unit cells of each of the one or more images classified for each category, selecting in each category an image according to a matching rate of the representative factor distribution data of the one or more images classified for each category and the representative factor distribution data of the unit cells of each of the one or more images classified for each category, and providing the selected image as a recommended target image,

selecting a pixel image for mosaicizing the determined target image based on tag-words of the plurality of images,

create a mosaic image for the target image based on the pixel image selected through the interface for selecting the pixel image.

12. The apparatus of claim 11 , wherein the processor is configured to:

use a deep learning neural network to select in each category the image and provide the recommended target image, and

acquire an input of a user of selecting the target image from the recommended target image and determine the target image among the plurality of images by providing a user interface.

13. The apparatus of claim 11 , wherein the processor is configured to:

detect in each category an image, in which the matching rate of the representative factor distribution data of the one or more images classified for each category and the representative factor distribution data of the unit cells of each of the one or more images classified for each category is equal or higher than a predetermined criterion.

14. The apparatus of claim 11 , wherein the processor is configured to:

provide an interface configured to receive selection of a tag-word or category from a user,

select one or more images which correspond to the selection of the tag-word or category as the pixel image, and

provide a recommended tag-word using a deep learning neural network.

15. The apparatus of claim 11 , wherein the processor is configured to divide the target image into the unit cells, match the pixel image for each unit cell, and insert the pixel image matched with the each unit cell to create the mosaic image.