Methods and Systems for Detecting and Assigning Attributes to Objects of Interest in Geospatial Imagery

Abstract

Exemplary systems and methods perform post-processing operations on computer vision model detections of objects of interest in geospatial imagery to detect and assign attributes to the detected objects of interest. For example, an exemplary post-processing system correlates multiple detections, made by a computer vision model, of an object of interest depicted within a set of images of a geospatial location, determines, based on the correlated detections, an attribute of the object of interest depicted within the set of images of the geospatial location, and selects the attribute for inclusion in a dataset for the object of interest. Corresponding methods and systems are also disclosed.

What is claimed is: 
     
         1 . A method comprising:
 correlating, by a post-processing system based on geospatial data associated with a set of images of a geospatial location, multiple detections of an object of interest depicted within the set of images of the geospatial location, wherein the correlating of the multiple detections comprises selecting the multiple detections for inclusion in a correlated set of detections from a plurality of possible detections of the object of interest;   determining, by the post-processing system based on the correlated detections, an attribute of the object of interest depicted within the set of images of the geospatial location; and   selecting, by the post-processing system, the attribute for inclusion in a dataset for the object of interest.   
     
     
         2 . The method of  claim 1 , wherein the set of images includes multiple images captured from different camera viewpoints of the geospatial location. 
     
     
         3 . The method of  claim 2 , wherein the correlating of the multiple detections further comprises:
 determining that the multiple detections intersect at a geospatial intersection; and   selecting the multiple detections for inclusion in the correlated set of detections based on the multiple detections intersecting at the geospatial intersection.   
     
     
         4 . The method of  claim 3 , wherein the determining that the multiple detections intersect at the geospatial intersection comprises performing one or more triangulation operations to determine that the geospatial intersection corresponds to each of the multiple detections. 
     
     
         5 . The method of  claim 3 , wherein the determining the attribute of the object of interest based on the correlated detections comprises:
 determining a geospatial position of the object of interest based on the geospatial intersection at which the correlated detections intersect.   
     
     
         6 . The method of  claim 1 , wherein the determining the attribute of the object of interest based on the correlated detections comprises:
 determining a consistency or an inconsistency across the correlated detections; and   determining the attribute of the object of interest based on the consistency or the inconsistency across the correlated detections.   
     
     
         7 . The method of  claim 6 , wherein the consistency or the inconsistency across the correlated detections comprises a consistency or an inconsistency in at least one of an object class and an object size across the correlated detections. 
     
     
         8 . The method of  claim 1 , wherein the determining the attribute of the object of interest based on the correlated detections comprises:
 determining a consistency or an inconsistency between a detection included in the correlated detections and geospatial data accessed from data source other than the set of images of the geospatial location; and   determining the attribute of the object of interest based on the consistency or the inconsistency between the detection and the geospatial data accessed from the data source other than the set of images of the geospatial location.   
     
     
         9 . A system comprising:
 at least one physical computing device configured to:
 correlate, based on geospatial data associated with a set of images of a geospatial location, multiple detections of an object of interest depicted within the set of images of the geospatial location, wherein the correlation of the multiple detections comprises selecting the multiple detections for inclusion in a correlated set of detections from a plurality of possible detections of the object of interest; 
 determine, based on the correlated detections, an attribute of the object of interest depicted within the set of images of the geospatial location; and 
 select the attribute for inclusion in a dataset for the object of interest. 
   
     
     
         10 . The system of  claim 9 , wherein the set of images includes multiple images captured from different camera viewpoints of the geospatial location. 
     
     
         11 . The system of  claim 10 , wherein the correlation of the multiple detections further comprises:
 the at least one physical computing device is further configured to:
 determine that the multiple detections intersect at a geospatial intersection; and 
 select the multiple detections for inclusion in the correlated set of detections based on the multiple detections intersecting at the geospatial intersection. 
   
     
     
         12 . The system of  claim 11 , wherein the determination of the attribute of the object of interest based on the correlated detections comprises:
 the at least one physical computing device is further configured to:
 determine a geospatial position of the object of interest based on the geospatial intersection at which the correlated detections intersect. 
   
     
     
         13 . The system of  claim 9 , wherein the determination of the attribute of the object of interest based on the correlated detections comprises:
 the at least one physical computing device is further configured to:
 determine a consistency or an inconsistency across the correlated detections; and 
 determine the attribute of the object of interest based on the consistency or the inconsistency across the correlated detections. 
   
     
     
         14 . The system of  claim 9 , the determination of the attribute of the object of interest based on the correlated detections comprises:
 the at least one physical computing device is further configured to:
 determine a consistency or an inconsistency between a detection included in the correlated detections and geospatial data accessed from data source other than the set of images of the geospatial location; and 
 determine the attribute of the object of interest based on the consistency or the inconsistency between the detection and the geospatial data accessed from the data source other than the set of images of the geospatial location. 
   
     
     
         15 . A non-transitory computer-readable medium storing instructions that, when executed, direct at least one processor of a computing device to:
 correlate, based on geospatial data associated with a set of images of a geospatial location, multiple detections of an object of interest depicted within the set of images of the geospatial location, wherein the correlation of the multiple detections comprises selecting the multiple detections for inclusion in a correlated set of detections from a plurality of possible detections of the object of interest;   determine, based on the correlated detections, an attribute of the object of interest depicted within the set of images of the geospatial location; and   select the attribute for inclusion in a dataset for the object of interest.   
     
     
         16 . The computer-readable medium of  claim 15 , wherein the set of images includes multiple images captured from different camera viewpoints of the geospatial location. 
     
     
         17 . The computer-readable medium of  claim 16 , wherein the correlation of the multiple detections further comprises:
 instructions, when executed, direct the at least one processor of the computing device to:
 determine that the multiple detections intersect at a geospatial intersection; and 
 select the multiple detections for inclusion in the correlated set of detections based on the multiple detections intersecting at the geospatial intersection. 
   
     
     
         18 . The computer-readable medium of  claim 17 , wherein the determination of the attribute of the object of interest based on the correlated detections comprises:
 instructions, when executed, direct the at least one processor of the computing device to:
 determine a geospatial position of the object of interest based on the geospatial intersection at which the correlated detections intersect. 
   
     
     
         19 . The computer-readable medium of  claim 15 , wherein the determination of the attribute of the object of interest based on the correlated detections comprises:
 instructions, when executed, direct the at least one processor of the computing device to:
 determine a consistency or an inconsistency across the correlated detections; and 
 determine the attribute of the object of interest based on the consistency or the inconsistency across the correlated detections. 
   
     
     
         20 . The computer-readable medium of  claim 19 , the determination of the attribute of the object of interest based on the correlated detections comprises:
 instructions, when executed, direct the at least one processor of the computing device to:
 determine a consistency or an inconsistency between a detection included in the correlated detections and geospatial data accessed from data source other than the set of images of the geospatial location; and 
 determine the attribute of the object of interest based on the consistency or the inconsistency between the detection and the geospatial data accessed from the data source other than the set of images of the geospatial location.