Original image generation system

• US 10,255,703 B2
• Filed: 12/18/2015
• Issued: 04/09/2019
• Est. Priority Date: 12/18/2015
• Status: Active Grant

• Current Assignee: eBay Inc.
• Original Assignee: eBay Inc.
• Inventors: Gonzales, Jr., Sergio Pinzon
• Primary Examiner(s): Johnson, Motilewa Good
• Application Number:
  US14/973,962
• Publication Number:
  US 20170178371A1
• Time in Patent Office: 1,208 Days
• Field of Search: None
• CPC Class Codes: G06F 18/24 Classification techniques
• CPC Class Codes: G06F 18/28 Determining representative reference patterns, e.g. by averaging or distorting; Generating dictionaries
• CPC Class Codes: G06T 11/00 2D [Two Dimensional] image generation
• CPC Class Codes: G06T 11/60 Editing figures and text; Combining figures or text
• CPC Class Codes: G06V 10/56 relating to colour
• CPC Class Codes: G06V 10/764 using classification, e.g. of video objects
• CPC Class Codes: G06V 10/772 Determining representative reference patterns, e.g. averaging or distorting patterns; Generating dictionaries

Abstract

A system and method for generating an original image are provided. In example embodiments, a user may select a category with a plurality of images. Common image attributes from the plurality of images within the user selected category is identified. A base image using the plurality of images associated with the identified common image attributes is generated. An original image is generated by varying attributes within the base image.

Claims

1. A system comprising:

one or more hardware processors and a memory including instructions that, when executed by the one or more hardware processors, causes the one or more hardware processors to perform operations comprising;

identifying a template for a category of products, the template defining locations of attributes of products belonging to the category;

identifying common attributes of products of the category based on the template;

selecting a first plurality of images from a second larger plurality of images, the first plurality of images selected based on their representation of products having the common attributes;

averaging corresponding pixel values across the selected images representing the products having the common attributes;

generating an overlaid image representing a new product based on the averaged pixel values; and

generating an original image by varying attributes of the new product based on the locations of the attributes defined by the template.

2. The system of claim 1, the operations further comprising altering a visual appearance of an attribute of an object represented by the overlaid image.

3. The system of claim 1, wherein:

the varying attributes includes changing a proportion of at least one common attribute of an object represented in the overlaid image.

4. The system of claim 3, wherein:

the changing of the proportion of at least one attribute includes at least one of elongating, shortening, widening, or slimming a specific portion of the at least one attribute of the object represented by the overlaid image.

5. The system of claim 1, the operations further comprising:

modifying at least one attribute of an object depicted in the overlaid image.

6. The system of claim 1, the operations further comprising:

calculating a similarity value between the original image and each of the plurality of images; and

varying attributes of an object represented by the original image until the similarity values are below a threshold.

7. The system of claim 1, the operations further comprising eliminating backgrounds from the plurality of images;

generating the overlaid image by overlaying the plurality of background-less images;

identifying areas from the overlaid image having a density of image information above a predetermined threshold; and

generating the original image from the identified areas.

8. The system of claim 7, the operations further comprising excluding areas of the overlaid image having a density of image information below the predetermined threshold from the original image.

9. The system of claim 8, the operations further comprising increasing an area of image overlap by iteratively adjusting an orientation of one or more of the background-less images and overlaying the adjusted images, and generating the base image from the adjusted images.

10. The system of claim 9, the operations further comprising iteratively adjust the orientations until a highest area of image overlap is detected.

11. The system of claim 1, the operations further comprising

determine a view angle for objects represented by each of the second plurality of images;

and

select the plurality of images based on a common view angle of the plurality of images.

12. A method comprising:

identifying a template for a category of products, the template defining locations of attributes of products in the category;

identifying common attributes of products of the category based on the template;

selecting a first plurality of images from a second larger plurality of images, the first plurality of images selected due to their representation of products having the common attributes;

averaging corresponding pixel values across the selected images representing the products having the common attributes;

generating an overlaid image representing a new product based on the averaged pixel values, and

generating an original image by varying attributes of the new product based on the locations of the attributes defined by the template, the generating being performed by at least one processor of a machine.

13. The method of claim 12, wherein:

the template defines a location of a sleeve on a shirt, wherein the products in the category are shirts, and the varying of attributes of the new product includes varying a length of a sleeve of a new shirt generated based on the averaged pixel values, the altering of the length of the sleeve being based on a location of the sleeve indicated by the template.

14. The method of claim 12,

further comprising performing a similarity check by comparing the pixel values of the original image with each of the plurality of images.

15. The method of claim 14, wherein:

the performing the similarity check includes;

calculating a similarity value between the original image and each of the plurality of images; and

where the similarity value is above a similarity threshold, additional variation attributes are added to the generated original image.

16. The method of claim 12, wherein:

the generating of the overlaid image further includes;

differentiating an item within each of the plurality of images using edge detection,

generating an item image without the background for each of the plurality of images; and

wherein the generating the overlaid image using the generated item image without the background.

17. The method of claim 16, wherein the operations further comprise:

overlaying the item image of each of the plurality of images; and

extracting overlapping areas where the number of overlapped items are above a predefined percentage to generate the overlaid image.

18. The method of claim 16, wherein:

the generating of the overlaid image includes averaging the pixel value of the generated item image without the background.

19. A non-transitory machine-readable storage medium comprising instructions that, when executed by one or more processors of a machine, cause the machine to perform operations comprising:

identifying common attributes of products of the category based on the template;

selecting a first plurality of images from a second larger plurality of images, the first plurality of images selected due to their representation of products having the common attributes;

averaging corresponding pixel values across the selected images that represent the products having the common attributes;

generating an overlaid image representing a new product based on the averaged pixel values; and

generating an original image by varying attributes of the new product based on the locations of the attributes defined by the template.

20. The machine-readable medium of claim 19, the operations further comprising altering a visual appearance of an attribute of an object represented by the overlaid image.

21. The machine-readable medium of claim 19, wherein the operations further comprise:

calculating a similarity value between the original image and each of the plurality of images; and

varying attributes of an object represented by the original image until the similarity values are below a threshold.