Limits...
A Geospatial Semantic Enrichment and Query Service for Geotagged Photographs.

Ennis A, Nugent C, Morrow P, Chen L, Ioannidis G, Stan A, Rachev P - Sensors (Basel) (2015)

Bottom Line: Nevertheless, it still remains a challenge to discover the "right" information for the appropriate purpose.To achieve this we have developed and implemented a semantic geospatial data model by which a photograph can be enrich with geospatial metadata extracted from several geospatial data sources based on the raw low-level geo-metadata from a smartphone photograph.We present the details of our method and implementation for searching and querying the semantic geospatial metadata repository to enable a user or third party system to find the information they are looking for.

View Article: PubMed Central - PubMed

Affiliation: School of Computing and Mathematics, University of Ulster, Coleraine BT370QB, UK. ennis-a1@email.ulster.ac.uk.

ABSTRACT
With the increasing abundance of technologies and smart devices, equipped with a multitude of sensors for sensing the environment around them, information creation and consumption has now become effortless. This, in particular, is the case for photographs with vast amounts being created and shared every day. For example, at the time of this writing, Instagram users upload 70 million photographs a day. Nevertheless, it still remains a challenge to discover the "right" information for the appropriate purpose. This paper describes an approach to create semantic geospatial metadata for photographs, which can facilitate photograph search and discovery. To achieve this we have developed and implemented a semantic geospatial data model by which a photograph can be enrich with geospatial metadata extracted from several geospatial data sources based on the raw low-level geo-metadata from a smartphone photograph. We present the details of our method and implementation for searching and querying the semantic geospatial metadata repository to enable a user or third party system to find the information they are looking for.

No MeSH data available.


Diagram showing MediaPlace architecture with the external datasets on the left and our photograph enrichment system on the right. It shows each component of the system and how each component interconnects.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4541944&req=5

sensors-15-17470-f001: Diagram showing MediaPlace architecture with the external datasets on the left and our photograph enrichment system on the right. It shows each component of the system and how each component interconnects.

Mentions: For the data model to handle all the criteria described above we choose to model the data in an ontology. This has the added benefit of being able to infer further relationships and also to reason over the data when executing a query. To handle the data model, the MediaPlace architecture was modified, as shown in Figure 1, so that in the POI Handler layer contains the RDF Enrichment Model Generator component, which creates the sets of triples representing the extracted objects and their attributes, such as latitude, longitude, place name, and place feature (building, statue, etc.) [14,15]. A POI is a significant location that may be of interest and, in relation to MediaPlace, is usually a building, statue, or business location. These triples represent subject-predicate-object. For example POIA hasName Belfast City Hall and Photograph isLookingAt POIA. During the semantic metadata repository population phase, predefined attributes are calculated, such as the distance to each POI from the photograph. This pre-calculation enables faster processing in our semantic geospatial metadata repository later on, particularly when searching. SWRL rules are also then run against the data stored in the semantic metadata repository to determine further information through inferences for example what POIs the photograph is looking at or the direction relationship each POI has to the photograph. This then enables searching, such as show all POIs to the south of a given photograph, or show all photographs to the east of a given POI.


A Geospatial Semantic Enrichment and Query Service for Geotagged Photographs.

Ennis A, Nugent C, Morrow P, Chen L, Ioannidis G, Stan A, Rachev P - Sensors (Basel) (2015)

Diagram showing MediaPlace architecture with the external datasets on the left and our photograph enrichment system on the right. It shows each component of the system and how each component interconnects.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4541944&req=5

sensors-15-17470-f001: Diagram showing MediaPlace architecture with the external datasets on the left and our photograph enrichment system on the right. It shows each component of the system and how each component interconnects.
Mentions: For the data model to handle all the criteria described above we choose to model the data in an ontology. This has the added benefit of being able to infer further relationships and also to reason over the data when executing a query. To handle the data model, the MediaPlace architecture was modified, as shown in Figure 1, so that in the POI Handler layer contains the RDF Enrichment Model Generator component, which creates the sets of triples representing the extracted objects and their attributes, such as latitude, longitude, place name, and place feature (building, statue, etc.) [14,15]. A POI is a significant location that may be of interest and, in relation to MediaPlace, is usually a building, statue, or business location. These triples represent subject-predicate-object. For example POIA hasName Belfast City Hall and Photograph isLookingAt POIA. During the semantic metadata repository population phase, predefined attributes are calculated, such as the distance to each POI from the photograph. This pre-calculation enables faster processing in our semantic geospatial metadata repository later on, particularly when searching. SWRL rules are also then run against the data stored in the semantic metadata repository to determine further information through inferences for example what POIs the photograph is looking at or the direction relationship each POI has to the photograph. This then enables searching, such as show all POIs to the south of a given photograph, or show all photographs to the east of a given POI.

Bottom Line: Nevertheless, it still remains a challenge to discover the "right" information for the appropriate purpose.To achieve this we have developed and implemented a semantic geospatial data model by which a photograph can be enrich with geospatial metadata extracted from several geospatial data sources based on the raw low-level geo-metadata from a smartphone photograph.We present the details of our method and implementation for searching and querying the semantic geospatial metadata repository to enable a user or third party system to find the information they are looking for.

View Article: PubMed Central - PubMed

Affiliation: School of Computing and Mathematics, University of Ulster, Coleraine BT370QB, UK. ennis-a1@email.ulster.ac.uk.

ABSTRACT
With the increasing abundance of technologies and smart devices, equipped with a multitude of sensors for sensing the environment around them, information creation and consumption has now become effortless. This, in particular, is the case for photographs with vast amounts being created and shared every day. For example, at the time of this writing, Instagram users upload 70 million photographs a day. Nevertheless, it still remains a challenge to discover the "right" information for the appropriate purpose. This paper describes an approach to create semantic geospatial metadata for photographs, which can facilitate photograph search and discovery. To achieve this we have developed and implemented a semantic geospatial data model by which a photograph can be enrich with geospatial metadata extracted from several geospatial data sources based on the raw low-level geo-metadata from a smartphone photograph. We present the details of our method and implementation for searching and querying the semantic geospatial metadata repository to enable a user or third party system to find the information they are looking for.

No MeSH data available.