US 9,811,539 B2
Hierarchical spatial clustering of photographs
Marek Vanco, Sunnyvale, CA (US); and Wei Luo, Saratoga, CA (US)
Assigned to Google Inc., Mountain View, CA (US)
Filed by Google Inc., Mountain View, CA (US)
Filed on Apr. 26, 2013, as Appl. No. 13/871,920.
Claims priority of provisional application 61/639,060, filed on Apr. 26, 2012.
Prior Publication US 2015/0154228 A1, Jun. 4, 2015
Int. Cl. G06F 17/30 (2006.01)
CPC G06F 17/30274 (2013.01) 23 Claims
OG exemplary drawing
 
1. A computer-implemented method of organizing photographs for display on an on-line map, the method comprising:
accessing a plurality of photographs, each photograph associated with a map feature on the on-line map and a score indicating the photograph's suitability;
selecting, by one or more computers, a set of candidate photographs from among the plurality of photographs, responsive to the photographs' scores;
determining, by the one or more computers, a cluster of selected photographs from the set of candidate photographs, the cluster having a predetermined shape determined by a measure of distribution of the selected photographs in the cluster and a threshold size determined based on geographic distance among locations associated with the set of candidate photographs, the selected photographs being selected such that a resulting shape and a resulting size of the cluster with the selected photographs are within the predetermined shape and the threshold size of the cluster, respectively; wherein the cluster corresponds to a particular zoom level for the on-line map, wherein clustering of the selected photographs at one zoom level is not dependent on clustering at another zoom level; and wherein the predetermined shape is further determined based on a shape of the map feature; and
storing the cluster as associated with each of the selected photographs in the set.