Back to results
US 8,897,604 B2Granted
Image quality analysis for searches
AssigneeAlibaba Group Holding Limited
InventorsYu Deng, Ke Chen
Life of this patent
Estimating the term…
Abstract
Image analysis includes: determining, using one or more processors, an image quality score associated with an image, including: determining a foreground and a background in the image; calculating a set of one or more characteristic parameters of the image based on the determined foreground and background; calculating the image quality score based at least in part on the set of characteristic parameters, wherein calculating the image quality score comprises using an image quality computation model that has been pre-trained; and in response to a search query, generating a set of search results that includes a set the images, wherein inclusion of the images or ranking of the search results is based at least in part on image quality scores associated with the set of images.
Claims15
A method comprising:
- determining, using one or more processors, an image quality score associated with an image, including:
- determining a foreground and a background in the image;
- calculating a set of one or more characteristic parameters of the image based on the foreground and the background;
- and
- calculating the image quality score based at least in part on the set of characteristic parameters, wherein calculating the image quality score comprises using an image quality computation model that has been pre-trained;
- and
- in response to a search query, generating a set of search results that includes a set of images,
- wherein inclusion of the images or ranking of the search results is based at least in part on image quality scores associated with the set of images;
- wherein determining the foreground and the background in the image comprises:
- determining an outer frame zone of the image, wherein the outer frame zone is determined using a preset value from edges of the image;
- clustering pixels in the outer frame zone to obtain a plurality of color zones,
- wherein each color zone corresponds to a color type;
- determining seed zones;
- expanding the seed zones to obtain the background in the image;
- and
- determining the foreground;
- and
- wherein determining seed zones comprises:
- calculating an area ratio of each color zone to an area of the outer frame zone;
- determining a second plurality of color zones that exceed a threshold area ratio;
- determining a zone flatness of each color zone in the second plurality of color zones;
- and
- determining seed zones, wherein the seed zones comprise color zones from the second plurality of color zones, and wherein the seed zones exceed a threshold color flatness value.
20 more elements
A system comprising:
- a one or more processors to:
- determine an image quality score associated with an image, including to:
- determine a foreground and a background in the image;
- calculate a set of one or more characteristic parameters of the image based on the foreground and the background;
- calculate the image quality score based at least in part on the set of characteristic parameters, wherein calculating the image quality score comprises using an image quality computation model that has been pre-trained;
- and
- in response to a search query, generate a set of search results that includes a set of images, wherein inclusion of the images or ranking of the search results is based at least in part on image quality scores associated with the set of images;
- wherein determining the foreground and the background in the image comprises:
- determining an outer frame zone of the image, wherein the outer frame zone is determined using a preset value from edges of the image;
- clustering pixels in the outer frame zone to obtain a plurality of color zones, wherein each color zone corresponds to a color type;
- determining seed zones;
- expanding the seed zones to obtain the background in the image;
- and
- determining the foreground;
- and
- wherein determining seed zones comprises:
- calculating an area ratio of each color zone to an area of the outer frame zone;
- determining a second plurality of color zones that exceed a threshold area ratio;
- determining a zone flatness of each color zone in the second plurality of color zones;
- and
- determining seed zones, wherein the seed zones comprise color zones from the second plurality of color zones, and wherein the seed zones exceed a threshold color flatness value;
- and
- a memory coupled to the one or more processors and configured to provide the one or more processors with instructions.
20 more elements
A computer program product being embodied in a tangible non-transitory computer readable storage medium and comprising computer instructions for:
- determining, using one or more processors, an image quality score associated with an image, including:
- determining a foreground and a background in the image;
- calculating a set of one or more characteristic parameters of the image based on the foreground and the background;
- calculating the image quality score based at least in part on the set of characteristic parameters, wherein calculating the image quality score comprises using an image quality computation model that has been pre-trained;
- and
- in response to a search query, generating a set of search results that includes a set of images, wherein inclusion of the images or ranking of the search results is based at least in part on image quality scores associated with the set of images;
- wherein determining the foreground and the background in the image comprises:
- determining an outer frame zone of the image, wherein the outer frame zone is determined using a preset value from edges of the image;
- clustering pixels in the outer frame zone to obtain a plurality of color zones,
- wherein each color zone corresponds to a color type;
- determining seed zones;
- expanding the seed zones to obtain the background in the image;
- and
- determining the foreground;
- and
- wherein determining seed zones comprises:
- calculating an area ratio of each color zone to an area of the outer frame zone;
- determining a second plurality of color zones that exceed a threshold area ratio;
- determining a zone flatness of each color zone in the second plurality of color zones;
- and
- determining seed zones, wherein the seed zones comprise color zones from the second plurality of color zones, and wherein the seed zones exceed a threshold color flatness value.
18 more elements
A method comprising:
- determining, using one or more processors, an image quality score associated with an image, including:
- determining a foreground and a background in the image;
- calculating a set of one or more characteristic parameters of the image based on the foreground and the background;
- and
- calculating the image quality score based at least in part on the set of characteristic parameters, wherein calculating the image quality score comprises using an image quality computation model that has been pre-trained;
- and
- in response to a search query, generating a set of search results that includes a set of images,
- wherein inclusion of the images or ranking of the search results is based at least in part on image quality scores associated with the set of images;
- wherein the characteristic parameters comprise a foreground-background contrast;
- and
- wherein foreground-background contrast is determined by:
- determining foreground color zones of the foreground and calculating an area ratio between the area of each foreground color zone and the area of the foreground;
- determining background color zones and calculating an area ratio between the area of each background color zone and the area of the foreground;
- calculating a Euclidean color-based distance between each foreground color zone and each background color zone;
- determining the foreground-background contrast based on the Euclidean color-based distances and the area ratios, wherein the foreground-background contrast is calculated using the formula:
- M
- =
- ∑
- i
- =
- 1
- ;
- j
- =
- 1
- n
- ,
- m
-
-
- LijAiDj
- ,
- wherein M is the foreground-background contrast, Lij is the Euclidean color-based distance between each foreground color zone and each background color zone, Ai is the area ratio of the foreground color zones to the area of the foreground and Dj is the area ratio of the background color zones to the area of the foreground.
30 more elements
A system comprising:
- a one or more processors to:
- determine an image quality score associated with an image, including:
- determine a foreground and a background in the image;
- calculate a set of one or more characteristic parameters of the image based on the foreground and the background;
- and
- calculate the image quality score based at least in part on the set of characteristic parameters, wherein calculating the image quality score comprises using an image quality computation model that has been pre-trained;
- and
- in response to a search query, generate a set of search results that includes a set of images,
- wherein inclusion of the images or ranking of the search results is based at least in part on image quality scores associated with the set of images;
- wherein the characteristic parameters comprise a foreground-background contrast;
- and
- wherein foreground-background contrast is determined by:
- determine foreground color zones of the foreground and calculate an area ratio between the area of each foreground color zone and the area of the foreground;
- determine background color zones and calculate an area ratio between the area of each background color zone and the area of the foreground;
- calculate a Euclidean color-based distance between each foreground color zone and each background color zone;
- and
- determine the foreground-background contrast based on the Euclidean color-based distances and the area ratios, wherein the foreground-background contrast is calculated using the formula:
- M
- =
- ∑
- i
- =
- 1
- ;
- j
- =
- 1
- n
- ,
- m
-
-
- LijAiDj
- ,
- wherein M is the foreground-background contrast, Lij is the Euclidean color-based distance between each foreground color zone and each background color zone, Ai is the area ratio of the foreground color zones to the area of the foreground and Dj is the area ratio of the background color zones to the area of the foreground;
- and
- a memory coupled to the one or more processors and configured to provide the one or more processors with instructions.
34 more elements
A computer program product being embodied in a tangible non-transitory computer readable storage medium and comprising computer instructions for:
- determining an image quality score associated with an image, including:
- determining a foreground and a background in the image;
- calculating a set of one or more characteristic parameters of the image based on the foreground and the background;
- and
- calculating the image quality score based at least in part on the set of characteristic parameters, wherein calculating the image quality score comprises using an image quality computation model that has been pre-trained;
- and
- in response to a search query, generating a set of search results that includes a set of images,
- wherein inclusion of the images or ranking of the search results is based at least in part on image quality scores associated with the set of images;
- wherein the characteristic parameters comprise a foreground-background contrast;
- and
- wherein foreground-background contrast is determined by:
- determining foreground color zones of the foreground and calculating an area ratio between the area of each foreground color zone and the area of the foreground;
- determining background color zones and calculating an area ratio between the area of each background color zone and the area of the foreground;
- calculating a Euclidean color-based distance between each foreground color zone and each background color zone;
- and
- determining the foreground-background contrast based on the Euclidean color-based distances and the area ratios, wherein the foreground-background contrast is calculated using the formula:
- M
- =
- ∑
- i
- =
- 1
- ;
- j
- =
- 1
- n
- ,
- m
-
-
- LijAiDj
- ,
- wherein M is the foreground-background contrast, Lij is the Euclidean color-based distance between each foreground color zone and each background color zone, Ai is the area ratio of the foreground color zones to the area of the foreground and Dj is the area ratio of the background color zones to the area of the foreground.
31 more elements
Dependencies are inferred from the claim text.