University of Twente Student Theses

Login

A semantic model for on-demand earth observation products

Akhtar, Anam (2021) A semantic model for on-demand earth observation products.

Link to full-text:https://ezproxy2.utwente.nl/login?url=https://library.itc.utwente.nl/login/2021/msc/gfm/akhtar.pdf
(only accessible for UT students and staff)
Abstract:Geoportals provide raw Earth Observation (EO) data which is subsequently processed to be used asEOproducts. One example of these products are spectral indices. Currently, geoportals provide only few spectral indices in their database although much more have been developed to serve various domains like water resources, urban planning, geology, etc. Decisions about which spectral indices to precompute is based on available storage and computing capacity as well as on previous demand. On-demand-computation can enable access to all the spectral indices without the need to precompute and store large amounts of EO products. For this purpose, comprehensive descriptions are needed that include all requirements and instructions to compute a spectral index. The research thesis identifies the different components required to build a semantic model for on-demand geoprocessing of spectral indices. For this, the mathematical formula for each spectral index is mentioned with their dependencies identified. This model has a vision of calculating the spectral indices as per user’s demand. To identify the user’s demand, competency questions are formulated. This further describes the various ways, the user can interact with the semantic model. As a result of competency questions, it was concluded that the model not only requires the formula but also the raw EO datasets which was populated as an ontology with the help of Open Geospatial Consortium (OGC):17-003r2 specification. One added functionality was that a user can also search for a spectral index via applications. This part is displayed in the EO4GEO BoK which was verified through a survey. This survey resulted in a positive response from the participants. Further, the ability of the model to extract the required product was validated by posing SPARQL queries to the formulated competency questions. The model was able to give the requested spectral index and their parameters as a result of user’s requirement. It is concluded that this model is successful in achieving its goal of describing the calculation of spectral indices explicitly.
Item Type:Essay (Master)
Faculty:ITC: Faculty of Geo-information Science and Earth Observation
Programme:Geoinformation Science and Earth Observation MSc (75014)
Link to this item:https://purl.utwente.nl/essays/88635
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page