Content collection search with robust content matching
View Patent ↗False positive match rates between query content and content in a collection may be reduced with a minimum content region test and/or a minimum features per scale test. The quality of correlations between query descriptors and content descriptors may be improved with a modified sub-region descriptor construction. Content regions associated with detected content features may be partitioned into disjoint sets of sub-regions that cover the content regions, the sub-regions modified so as to at least partially overlap, and descriptor components generated for the modified sub-regions. Matching of feature-sparse content may be improved by adding blurred versions to the collection.
1. A computer-implemented method for searching a content collection, comprising:
under control of one or more computer systems configured with executable instructions,
receiving a search request specifying query content to be matched with content in the content collection, the content collection characterized by a set of content descriptors;
determining a set of query descriptors for the query content, the query content including one or more regions, the one or more regions including a plurality of sub-regions, at least one of the set of query descriptors associated with query descriptor components corresponding to at least partially overlapping sub-regions of one of the one or more regions that includes an identified feature of the query content;
identifying a subset of the set of content descriptors that corresponds to at least a portion of the set of query descriptors; and
providing for presentation at least reference to a subset of the content collection selected based at least in part on the subset of the set of content descriptors.
2. A computer-implemented method according to claim 1 , wherein the query content comprises a digitized image, and the query content region is a set of pixels of the digitized image.
3. A computer-implemented method according to claim 1 , wherein said at least partially overlapping sub-regions of the query content region include at least one sub-region corresponding to at least one expanded version of a set of disjoint sub-regions that cover the query content region.
4. A computer-implemented method according to claim 3 , wherein said at least partially overlapping sub-regions cover the query content region.
5. A computer-implemented method according to claim 1 , wherein said at least partially overlapping sub-regions of the query content region include at least one sub-region corresponding to at least one contracted version of a set of disjoint sub-regions that cover the query content region.
6. A computer-implemented method for searching a content collection, comprising:
under control of one or more computer systems configured with executable instructions,
determining a set of content descriptors for the content collection, each content piece in the content collection including one or more regions, the one or more regions including a plurality of sub-regions, at least one of the set of content descriptors associated with content descriptor components corresponding to at least partially overlapping sub-regions of one of the one or more regions of a content piece in the content collection;
determining a set of query descriptors for query content of a search request;
identifying a subset of the set of content descriptors that corresponds to at least a portion of the set of query descriptors; and
providing for presentation at least reference to a subset of the content collection selected based at least in part on the subset of the set of content descriptors.
7. A computer-implemented method according to claim 6 , wherein said at least partially overlapping sub-regions cover the region of the content piece.
8. A computer-implemented method according to claim 7 , wherein each of the content descriptor components corresponds to one of said at least partially overlapping sub-regions.
9. A computer-implemented method according to claim 8 , wherein each of said at least partially overlapping sub-regions corresponds to a plurality of the content descriptor components.
10. A computer-implemented method according to claim 9 , wherein the region of the content piece is associated with a plurality of intensity gradients and the plurality of the content descriptor components characterize the plurality of intensity gradients within the corresponding sub-region.
11. A system for searching a content collection, comprising:
at least one device processor; and
memory storing instructions that, when executed by the at least one device processor, causes the system to:
determine a set of content descriptors for the content collection, each content piece in the content collection including one or more regions, the one or more regions including a plurality of sub-regions, at least one of the set of content descriptors being associated with content descriptor components corresponding to at least partially overlapping sub-regions of one of the one or more regions of a content piece in the content collection;
determine a set of query descriptors for query content of a search request;
identify a subset of the set of content descriptors that corresponds to at least a portion of the set of query descriptors; and
provide for presentation at least reference to a subset of the content collection selected based at least in part on the subset of the set of content descriptors.
12. A system according to claim 11 , wherein said at least partially overlapping sub-regions cover the region of the content piece.
13. A system according to claim 12 , wherein each of the content descriptor components corresponds to one of said at least partially overlapping sub-regions.
14. A system according to claim 13 , wherein each of said at least partially overlapping sub-regions corresponds to a plurality of the content descriptor components.
15. A system according to claim 14 , wherein the region of the content piece is associated with a plurality of intensity gradients and the plurality of the content descriptor components characterize the plurality of intensity gradients within the corresponding sub-region.
16. A non-transitory computer-readable storage medium including instructions that, when executed by at least one processor of a computing system, cause the computing system to:
determine a set of content descriptors for the content collection, each content piece in the content collection including one or more regions, the one or more regions including a plurality of sub-regions, at least one of the set of content descriptors being associated with content descriptor components corresponding to at least partially overlapping sub-regions of one of the one or more regions of a content piece in the content collection;
determine a set of query descriptors for query content of a search request;
identify a subset of the set of content descriptors that corresponds to at least a portion of the set of query descriptors; and
provide for presentation at least reference to a subset of the content collection selected based at least in part on the subset of the set of content descriptors.
17. A non-transitory computer-readable storage medium according to claim 16 , wherein said at least partially overlapping sub-regions cover the region of the content piece.
18. A non-transitory computer-readable storage medium according to claim 17 , wherein each of the content descriptor components corresponds to one of said at least partially overlapping sub-regions.
19. A non-transitory computer-readable storage medium according to claim 18 , wherein each of said at least partially overlapping sub-regions corresponds to a plurality of the content descriptor components.
20. A non-transitory computer-readable storage medium according to claim 19 , wherein the region of the content piece is associated with a plurality of intensity gradients and the plurality of the content descriptor components characterize the plurality of intensity gradients within the corresponding sub-region.