Image management through lexical representations

• US 20080088642A1
• Filed: 10/17/2006
• Published: 04/17/2008
• Est. Priority Date: 10/17/2006
• Status: Active Grant

• Current Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Original Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Inventors: Obrador, Pere, Wu, Peng
• Granted Patent: US 7,755,646 B2
• Time in Patent Office: Days
• US Class Current: 345/619
• 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

Abstract

In a method for image management, image data for an image containing objects is accessed. Graphical representations of the objects are generated and the centroids and the sizes of the graphical representations are determined. The locations of the centroids are determined and the morphologies of the graphical representations based upon the locations of the centroids are determined. Human readable lexical representations of the locations of the centroids, the sizes, the colors and morphologies of the graphical representations are assigned and are stored in a database, which is searchable through human readable lexicon.

Claims

1. A method for image management comprising:

accessing image data for an image containing objects;

generating graphical representations of the objects;

determining the centroids and the sizes of the graphical representations;

determining the locations of the centroids;

determining morphologies of the graphical representations based upon the determined locations of the centroids, wherein the morphologies of the graphical representations comprise physical relationships of the graphical representations with respect to each other;

assigning human readable lexical representations of the locations of the centroids, the sizes, the colors and morphologies of the graphical representations; and

storing the assigned human readable lexical representations in a database searchable through human readable lexicon.

2. The method according to claim 1, further comprising:

for images containing a plurality of graphical representations, determining a morphology of the plurality of graphical representations with respect to each other; and

assigning lexical representations of the relative positions of the plurality of graphical representations.

3. The method according to claim 2, wherein assigning lexical representations of the plurality of graphical representations further comprises assigning lexical representations pertaining to which of the graphical representations are in contact with which of the graphical representations and which of the graphical representations are in contact with at least one border of a respective image.

4. The method according to claim 3, wherein generating the graphical representations of the objects further comprises quantizing the image to identify graphical representations having at least one of a consistent and a common characteristic to identify the plurality of graphical representations.

5. The method according to claim 1, wherein storing the assigned human readable lexical representations in a database searchable through human readable lexicon further comprises storing the assigned human readable lexical representations in a metadata of the image data.

6. The method according to claim 1, wherein assigning human readable lexical representations of the sizes of the graphical representations further comprises assigning a quantized set of lexical size designations to the regions.

7. The method according to claim 1, wherein charting the locations of the centroids further comprises charting the locations of the centroids on a morpho-lexical histogram, dividing the morpho-lexical histogram into a plurality of virtual regions, wherein determining the morphologies of the graphical representations further comprises determining which of the graphical representations are contained within which of the plurality of virtual regions, and wherein assigning human readable lexical representations of the morphologies of the graphical representations further comprises assigning human readable lexical representations of the morphologies based upon the relationships between the virtual regions and the graphical representations contained within the virtual regions.

8. The method according to claim 7, wherein assigning human readable lexical representations of the morphologies further comprises assigning human readable position designations for each of the virtual regions to distinguish the locations of the virtual regions with respect to each other and to thereby identify the locations of the graphical representations with respect to each other.

9. The method according to claim 1, further comprising:

retrieving at least one image from the database in response to receipt of a search query, wherein the search query comprises a text-based query.

10. The method according to claim 9, further comprising:

determining one or more characteristics of the search query, wherein the one or more characteristics define relative locations of various graphical representations in a desired image.

11. The method according to claim 10, further comprising:

comparing the one or more characteristics with the assigned human readable lexical representations stored in the database;

determining whether one or more of the images stored in the database contains features that substantially match the one or more characteristics; and

wherein retrieving at least one image from the database further comprises retrieving at least one image that contains features that substantially match the one or more characteristics.

12. The method according to claim 11, further comprising:

ranking a plurality of images based on similarities between the one or more characteristics and the assigned human readable lexical representations of the images in response to a determination that a plurality of images contains features that substantially match the one or more characteristics.

13. The method according to claim 1, further comprising:

retrieving at least one image from the database in response to receipt of a search query, wherein the search query comprises a scanned image;

morpho-lexically processing the scanned image to derive lexical representations of the scanned image;

comparing the lexical representations of the scanned image with the assigned human readable lexical representations stored in the database;

determining whether one or more of the images stored in the database contains features that substantially match the lexical representations of the scanned image; and

wherein retrieving at least one image from the database further comprises retrieving at least one image that contains features that substantially match the lexical representations of the scanned image.

14. An image management system comprising:

processing circuitry configured to access image data of a plurality of images, said processing circuitry being configured to generate graphical representations of objects contained in the plurality of images, to determine the centroids and the sizes of the graphical representations, to determine morphologies of the graphical representations, and to assign human readable lexical representations of the locations of the centroids, the sizes, the colors and the morphologies of the graphical representations; and

storage circuitry configured to store the plurality of images and to store the plurality of human readable lexical representations of the plurality of images.

15. The image management system according to claim 14, wherein the processing circuitry is further configured to quantize the images to identify graphical representations in the respective images having at least one of a consistent and a common characteristic to thereby identify the plurality of graphical representations.

16. The image management system according to claim 14, wherein the processing circuitry is further configured to chart the locations of the centroids on a morpho-lexical histogram, to divide the morhpho-lexical histogram into a plurality of virtual regions, and to assign human readable representations of the morphologies of the graphical representations based upon the relationships between the virtual regions and the graphical representations contained within the virtual regions.

17. The image management system according to claim 14, wherein the processing circuitry is configured to retrieve at least one image from the storage circuitry in response to receipt of a search query, and wherein the processing circuitry is further configured to determine one or more characteristics of the search query, wherein the one or more characteristics define at least the relative locations of various graphical representations in a desired image.

18. The image management system according to claim 17, wherein the search query comprises a text-based query, and wherein the processing circuitry is further configured to analyze the text-based query to determine features relevant to the text-based query and to search for images in the storage circuitry that are similar to the determined features by comparing the determined features with the assigned human readable lexical representations of the images.

19. The image management system according to claim 17, wherein the search query comprises a scanned image, and wherein the processing circuitry is further configured to morpho-lexically process the scanned image to thereby derive lexical representations of the scanned image, and wherein the processing circuitry is still further configured to search for images in the storage circuitry that are similar to the lexical representations by comparing the lexical representations of the scanned image with the assigned human readable lexical representations.

20. A computer readable storage medium on which is embedded one or more computer programs, said one or more computer programs implementing a method for image management, said one or more computer programs comprising a set of instructions for:

accessing image data for a plurality of images containing objects;

generating graphical representations of the objects;

determining the centroids and the sizes of the graphical representations;

determining the locations of the centroids for each of the plurality of images;

determining morphologies of the graphical representations based upon the charted locations of the centroids, wherein the morphologies of the graphical representations comprise physical relationships of the graphical representations with respect to each other;

assigning human readable lexical representations of the locations of the centroids, the sizes, the colors and morphologies of the graphical representations for each of the plurality of images; and

storing the assigned human readable lexical representations in a database, said database being searchable through search queries containing human readable lexicon.