Automatically selecting and superimposing images for aesthetically pleasing photo creations

Patent metadata

Field	Value
Publication number	US-10810775-B2
Application number	US-201916280267-A
Country	US
Kind code	B2
Filing date	Feb 20, 2019
Priority date	Feb 20, 2019
Publication date	Oct 20, 2020
Grant date	Oct 20, 2020

Abstract

This application relates generally to computer-implemented methods and systems for computer graphics processing. Specifically, the application involves automatically selecting and blending images based on image aesthetic scores. Computer systems are provided that may blend an automatically selected foreground image with an automatically selected background image to produce an editable photo creation.

First claim

The invention claimed is:

1. A computer-implemented method that includes one or more processors performing operations comprising:
receiving a first plurality of images associated with a user;
generating, for a first image from the first plurality of images, a face quality score based on a number of faces in the first image and a face prominence ratio determined from a diameter of a face in the first image;
selecting, based on the face quality score, the first image;
identifying, from a second plurality of images associated with the user, a second image, wherein the second image depicts zero human faces; and
blending the first image with the second image to produce a photo creation.

2. The computer-implemented method of claim 1 , the operations further comprising:
identifying, based on an image aesthetic score of each image within the second plurality of images, the second image.

3. The computer-implemented method of claim 1 , wherein the second image is associated with an image tag that indicates contents within an image.

4. The computer-implemented method of claim 1 , wherein the blending of the first image and the second image comprises:
inverting one or more pixels of the first image to generate a first set of inverted pixels;
inverting one or more pixels of the second image to generate a second set of inverted pixels; and
multiplying the first set of inverted pixels with the second set of inverted pixels.

5. The computer-implemented method of claim 4 , the operations further comprising inverting a product of the first set of inverted pixels and the second set of inverted pixels.

6. The computer-implemented method of claim 1 , the operations further comprising:
generating a first image layer based on the first image;
generating a second image layer comprising a masked version of the first image;
generating, based on the second image layer, a third image layer comprising photometric parameters associated with the second image layer;
generating a fourth image layer based on the second image;
creating the photo creation with a double exposure effect by at least blending pixels of the second image layer and the fourth image layer; and
outputting the photo creation with the double exposure effect with modifiable layers as an output file.

7. The computer-implemented method of claim 6 , the operations further comprising:
subsequent to the photo creation being created, modifying the first image layer, the second image layer, the third image layer, or the fourth image layer.

8. A non-transitory computer-readable storage medium having stored thereon instructions that are executable by one or more processing devices to perform operations comprising:
receiving a first plurality of images associated with a user;
generating, for a first image from the first plurality of images, a face quality score based on a number of faces in the first image and a face prominence ratio determined from a diameter of a face in the first image;
selecting, based on the face quality score, the first image;
identifying, from a second plurality of images associated with the user a second image, wherein the second image depicts zero human faces; and
blending the first image with the second image to produce a photo creation.

9. The non-transitory computer-readable storage medium of claim 8 , the operations further comprising:
identifying, based on an image aesthetic score of each image within the second plurality of images, the second image.

10. The non-transitory computer-readable storage medium of claim 8 , wherein the second image is associated with an image tag that indicates contents within an image.

11. The non-transitory computer-readable storage medium of claim 8 , wherein blending of the first image and the second image comprises:
inverting one or more pixels of the first image to generate a first set of inverted pixels;
inverting one or more pixels of the second image to generate a second set of inverted pixels; and
multiplying the first set of inverted pixels with the second set of inverted pixels.

12. The non-transitory computer-readable storage medium of claim 11 , the operations further comprising inverting a product of the first set of inverted pixels and the second set of inverted pixels.

13. The non-transitory computer-readable storage medium of claim 12 , the operations further comprising:
generating a first image layer based on the first image;
generating a second image layer comprising a masked version of the first image;
generating, based on the second image layer, a third image layer comprising photometric parameters associated with the second image layer;
generating a fourth image layer based on the second image;
creating the photo creation with a double exposure effect by at least blending pixels of the second image layer and the fourth image layer; and
outputting the photo creation with the double exposure effect with modifiable layers as an output file.

14. The non-transitory computer-readable storage medium of claim 13 , the operations further comprising:
subsequent to the photo creation being created, modifying the first image layer, the second image layer, the third image layer, or the fourth image layer.

15. A system comprising:
one or more processors; and
a memory coupled with the one or more processors, the memory configured to store instructions that when executed by the one or more processors cause the one or more processors to:
receive a first plurality of images associated with a user;
generate, for a first image from the first plurality of images, a face quality score based on a number of faces in the first image and a face prominence ratio determined from a diameter of a face in the first image;
select, based on the face quality score, the first image;
identify, from a second plurality of images associated with the user, a second image, wherein the second image depicts zero human faces; and
blend the first image with the second image to produce a photo creation.

16. The system of claim 15 , wherein the instructions when executed by the one or more processors, further cause the one or more processors to identify, based on an image aesthetic score of each image within the second plurality of images, the second image.

17. The system of claim 15 , wherein the second image is associated with an image tag that indicates contents within an image.

18. The system of claim 15 , wherein the instructions when executed by the one or more processors, further cause the one or more processors to:
invert one or more pixels of the first image to generate a first set of inverted pixels;
invert one or more pixels of the second image to generate a second set of inverted pixels; and
multiply the first set of inverted pixels with the second set of inverted pixels.

19. The system of claim 18 , wherein the instructions, when executed by the one or more processors, further cause the one or more processors to invert a product of the first set of inverted pixels and the second set of inverted pixels.

20. The system of claim 15 , wherein the instructions that when executed by the one or more processors further cause the one or more processors to:
generate a first image layer based on the first image;
generate a second image layer comprising a masked version of the first image;
generate, based on the second image layer, a third image layer comprising photometric parameters associated with the second image layer;
generate a fourth image layer based on the second image;

Assignees

• Adobe Inc

Inventors

• Gupta Sonali
• Bhatt Sameer
• Somani Aanchal

Classifications

• G06V40/161

  Detection; Localisation; Normalisation · CPC title

• G06V10/82

  using neural networks · CPC title

• G06V10/764

  using classification, e.g. of video objects · CPC title

• G06V40/165

  using facial parts and geometric relationships · CPC title

• G06T11/60 Primary

  Creating or editing images; Combining images with text · CPC title