IMAGE QUALITY ASSESSMENT TO MERCHANDISE AN ITEM

• US 20120128239A1
• Filed: 11/18/2011
• Published: 05/24/2012
• Est. Priority Date: 11/18/2010
• Status: Active Grant

• Current Assignee: eBay Inc.
• Original Assignee: eBay Inc.
• Inventors: Chittar, Naren, Dasdan, Ali, Ghatare, Sanjay Pundlkrao, Goswami, Anjan, Gaikwad, Sandip Namdeo, Chung, Sung Hwan
• Granted Patent: US 8,675,957 B2
• Time in Patent Office: Days
• US Class Current: 382/162
• CPC Class Codes: G06F 16/24578 using ranking
• CPC Class Codes: G06F 16/58 Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
• CPC Class Codes: G06F 16/5838 using colour
• CPC Class Codes: G06F 16/5854 using shape and object relationship
• CPC Class Codes: G06Q 30/0243 Comparative campaigns
• CPC Class Codes: G06Q 30/0276 Advertisement creation
• CPC Class Codes: G06T 2207/30168 Image quality inspection
• CPC Class Codes: G06T 7/0002 Inspection of images, e.g. flaw detection

Abstract

Image-based features may be significantly correlated with click-through rates of images that depict a product, which may provide a more formal basis for the informal notion that good quality images will result in better click-through rates, as compared to poor quality images. Accordingly, an image assessment machine is configured to analyze image-based features to improve click-through rates for shopping search applications (e.g., a product search engine). Moreover, the image assessment machine may rank search results based on image quality factors and may notify sellers about low quality images. This may have the effect of improving the brand value for an online shopping website and accordingly have a positive long-term impact on the online shopping website.

Claims

1. A method comprising:

accessing an image that depicts an item,

the image having a plurality of visual characteristics that are quantifiable by an analysis of image data included in the image;

performing the analysis of the image data included in the image that depicts the item and that has the plurality of visual characteristics;

determining a score that quantifies a visual characteristic among the plurality of visual characteristics of the image that depicts the item,

the determining of the score being based on the analysis of the image data included in the image that depicts the item,

the determining of the score being performed by a processor of a machine; and

performing an operation that makes reference to the image,

the operation being performed based on the score that quantifies the visual characteristic among the plurality of visual characteristics of the image that depicts the item.

2. The method of claim 1 further comprising:

receiving the image as a submission from a seller of the item for inclusion in an advertisement of the item; and

wherein

the performing of the operation includes providing a warning to the seller that the image is of low quality.

3. The method of claim 1 further comprising:

receiving the image as a submission from a seller of the item for inclusion in an advertisement of the item; and

wherein

the performing of the operation includes providing a suggestion to the seller that the received image be replaced with an alternative image of higher quality.

4. The method of claim 1, wherein:

the visual characteristic is a brightness of the image;

the performing of the analysis of the image data includes analyzing the brightness of the image;

the score quantifies the brightness of the image; and

the performing of the operation is based on the brightness of the image.

5. The method of claim 1, wherein:

the visual characteristic is a contrast of the image;

the performing of the analysis of the image data includes analyzing the contrast of the image;

the score quantifies the contrast of the image; and

the performing of the operation is based on the contrast of the image.

6. The method of claim 1, wherein:

the visual characteristic is a saturation of the image;

the performing of the analysis of the image data includes analyzing the saturation of the image;

the score quantifies the saturation of the image; and

the performing of the operation is based on the saturation of the image.

7. The method of claim 1, wherein:

the visual characteristic is a colorfulness of the image;

the performing of the analysis of the image data includes analyzing the colorfulness of the image;

the score quantifies the colorfulness of the image; and

the performing of the operation is based on the colorfulness of the image.

8. The method of claim 1, wherein:

the visual characteristic is a dynamic range of the image;

the performing of the analysis of the image data includes analyzing the dynamic range of the image;

the score quantifies the dynamic range of the image; and

the performing of the operation is based on the dynamic range of the image.

9. The method of claim 1, wherein:

the visual characteristic is a size of the image;

the performing of the analysis of the image data includes analyzing the size of the image;

the score quantifies the size of the image; and

the performing of the operation is based on the size of the image.

10. The method of claim 1, wherein:

the visual characteristic is an aspect ratio of the image;

the performing of the analysis of the image data includes analyzing the aspect ratio of the image;

the score quantifies the aspect ratio of the image; and

the performing of the operation is based on the aspect ratio of the image.

11. The method of claim 1 further comprising:

segmenting the image into a foreground portion of the image that depicts the item and a background portion of the image that depicts no part of the item;

wherein

the visual characteristic is a uniformity of the foreground portion of the image;

the performing of the analysis of the image data includes analyzing the uniformity of the foreground portion of the image;

the score quantifies the uniformity of the foreground portion of the image; and

the performing of the operation is based on the uniformity of the foreground portion of the image.

12. The method of claim 1 further comprising:

segmenting the image into a foreground portion of the image that depicts the item and a background portion of the image that depicts no part of the item;

wherein

the visual characteristic is a uniformity of the background portion of the image;

the performing of the analysis of the image data includes analyzing the uniformity of the background portion of the image;

the score quantifies the uniformity of the background portion of the image; and

the performing of the operation is based on the uniformity of the background portion of the image.

13. The method of claim 1 further comprising:

segmenting the image into a foreground portion of the image that depicts the item and a background portion of the image that depicts no part of the item;

wherein

the visual characteristic is a ratio of a number of foreground pixels from the foreground portion to a number of background pixels from the background portion;

the performing of the analysis of the image data includes analyzing the ratio of the number of foreground pixels to the number of background pixels;

the score quantifies the ratio of the number of foreground pixels to the number of background pixels; and

the performing of the operation is based on the ratio of the number of foreground pixels to the number of background pixels.

14. The method of claim 1 further comprising:

determining a plurality of scores as feature vectors that quantify the plurality of visual characteristics of the image; and

determining an overall score representative of an overall visual quality of the image based on the plurality of scores;

wherein

the performing of the operation is based on the overall score that represents the overall visual quality of the image that depicts the item.

15. The method of claim 14, wherein:

the overall score represents an average brightness of the image; and

the performing of the operation is based on the average brightness of the image.

16. The method of claim 14, wherein:

the overall score represents an average contrast to the image; and

the performing of the operation is based on the average contrast to the image.

17. The method of claim 14, wherein:

the overall score represents an average lightness of the image,

the average lightness indicating an average distance from white in a color space; and

the performing of the operation is based on the average lightness of the image.

18. The method of claim 14 further comprising:

segmenting the image into a foreground portion of the image that depicts the item and a background portion of the image that depicts no part of the item;

wherein

the overall score represents an average lightness of the background portion of the image,

the average lightness indicating an average distance from white in a color space; and

the performing of the operation is based on the average lightness of the background portion of the image.

19. The method of claim 14 further comprising:

segmenting the image into a foreground portion of the image that depicts the item and a background portion of the image that depicts no part of the item;

wherein

the overall score represents an average uniformity of the background portion of the image; and

the performing of the operation is based on the average uniformity of the background portion of the image.

20. The method of claim 1 further comprising:

determining a shape of the item by segmenting the image;

accessing a database that correlates the shape of the item with textual information that describes the item; and

wherein

the performing of the operation includes providing the textual information based on the shape of the item being correlated with the textual information.

21. The method of claim 20 further comprising:

receiving different textual information as a submission for a description of the item from a seller of the item; and

wherein

the performing of the operation includes providing a warning to the seller that the different textual information does not describe the item.

22. The method of claim 20 further comprising:

receiving different textual information as a submission for a description of the item from a seller of the item; and

wherein

the performing of the operation includes providing a suggestion to the seller that the item be described by the textual information accessed from the database.

23. The method of claim 1, wherein:

the performing of the operation includes determining a rank of the image of the item and providing the image of the item within a search result,

the search result being provided according to the determined rank among a plurality of search results.

24. The method of claim 23 further comprising:

determining a click-through rate of an advertisement that includes the image; and

wherein

the determining of the rank of the image is based on the click-through rate of the advertisement.

25. The method of claim 1, wherein:

the image depicts the item under a first lighting condition;

the performing of the analysis of the image data includes;

generating a histogram of the image,

partitioning energy values from the histogram of the image into multiple partitions of the histogram, and

determining a first energy value of a partition among the multiple partitions based on a second energy value of an adjacent partition among the multiple partitions; and

the performing of the operation includes providing an indication that the item depicted in the image is also depicted in a further image that depicts the item under a second lighting condition.

26. 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:

accessing an image that depicts an item,

the image having a plurality of visual characteristics that are quantifiable by an analysis of image data included in the image;

performing the analysis of the image data included in the image that depicts the item and that has the plurality of visual characteristics;

determining a score that quantifies a visual characteristic among the plurality of visual characteristics of the image that depicts the item,

the determining of the score being based on the analysis of the image data included in the image that depicts the item,

the determining of the score being performed by the one or more processors of the machine; and

performing an operation that makes reference to the image,

the operation being performed based on the score that quantifies the visual characteristic among the plurality of visual characteristics of the image that depicts the item.

27. The non-transitory machine-readable storage medium of claim 26, wherein the operations further comprise:

receiving the image as a submission from a seller of the item for inclusion in an advertisement of the item; and

wherein

the performing of the operation includes providing a warning to the seller that the image is of low quality.

28. A system comprising:

an access module configured to access an image that depicts an item,

the image having a plurality of visual characteristics that are quantifiable by an analysis of image data included in the image;

an analysis module configured to perform the analysis of the image data included in the image that depicts the item and that has the plurality of visual characteristics;

a processor configured by a score module that configures the processor to determine a score that quantifies a visual characteristic among the plurality of visual characteristics of the image that depicts the item,

the determining of the score being based on the analysis of the image data included in the image that depicts the item,

the determining of the score being performed by a processor of a machine; and

an action module configured to perform an operation that makes reference to the image,

the operation being performed based on the score that quantifies the visual characteristic among the plurality of visual characteristics of the image that depicts the item.

29. The system of claim 28, wherein:

the access module is configured to receive the image as a submission from a seller of the item for inclusion in an advertisement of the item; and

wherein

the operation module is configured to provide a warning to the seller that the image is of low quality.