Computerized geo-referencing for images

A computing system, comprising:

• a hardware processor connected to memory;
• and
• a geo-referencing module stored on a non-transitory computer readable medium and configured with instructions that when executed by the processor cause the processor to:
• extract features from an image of a location as extracted features, wherein the image includes geo-reference metadata describing geo-locations of objects depicted within the image;
• determine similarities between the extracted features and known features within vector data for the location;
• generate a data structure representing similarities between the extracted features and the known features that exceed a similarity threshold, wherein the data structure is generated to include at least a graph representing edge connections between one or more of a first set of nodes representing the extracted features and one or more of a second set of nodes representing the known features, wherein the edges are created between the nodes in the first set of nodes and the nodes in the second set of nodes based upon the similarities that exceed the similarity threshold;
• scale the extracted features to create extracted features with same total lengths;
• for an extracted feature, divide lengths of edges of the extracted feature by a total length of the extracted feature;
• process the similarities within the data structure to identify a match between a first extracted feature and a first known feature based upon the first extracted feature having a highest similarity with the first known feature;
• assign geo-location data from the vector data for the first known feature to the first extracted feature;
• identify a position of an object represented by the first extracted feature within the location using the geo-location data;
• and
• wherein when the image does not comprise geo-reference metadata, geo-location data is determined, generated and assigned as new geo-reference metadata for the image.

A computer-implemented method, the computer-implemented method involving a computing device comprising a processor, and the computer-implemented method comprising:

• extracting, by the processor, features from an image of a location as extracted features, wherein the image comprises geo-reference metadata describing geo-locations of objects depicted within the image;
• determining, by the processor, similarities between the extracted features and known features within vector data for the location;
• generating, by the processor, a data structure representing similarities between the extracted features and the known features that exceed a similarity threshold, wherein the data structure is generated by at least generating a graph representing edge connections between one or more of a first set of nodes representing the extracted features and one or more of a second set of nodes representing the known features, wherein the edges are created between the nodes in the first set of nodes and the nodes in the second set of nodes based upon the similarities that exceed the similarity threshold;
• scaling the extracted features to create extracted features with same total lengths;
• for an extracted feature, dividing lengths of edges of the extracted feature by a total length of the extracted feature;
• processing, by the processor, the similarities to identify a match between a first extracted feature and a first known feature based upon the first extracted feature having a highest similarity with the first known feature;
• modifying, by the processor, the geo-reference metadata for the first extracted feature based upon geo-location data of the first known feature;
• and
• wherein when the image does not comprise geo-reference metadata, geo-location data is determined, generated and assigned as new geo-reference metadata for the image.