| With rapid progresses in aerospace engineering and remote sensing technology, acquisition of remote sensing images shows a growing trend toward multi-platform, multi-sensor, multi-angle, high spectral resolution and high spatio-temporal resolution. Nowadays, the burgeoning remote-sensing data have been widely applied to many related domains, such as resources survey, environmental protection, disaster prevention and reduction, precision agriculture, forestry industry, water conservation, urban planning and management. Due to the absence of sharing mechanism, the conflicts between rapid acquisition for remote sensing data and actual demands are being increasingly intensified. Service-oriented architecture provides a possible pathway for clear up the conflicts, which is expected to offer a high-efficiency, cooperative support for practical tasks by using the numerous remote-sensing data and the corresponding processing functions distributed on internet. The existing technologies and approaches for geo-spatial information services, however, can only realize sharing and interoperability of geo-spatial information services at the syntax level, which is far less sufficient to meet the need for applications. Thus, it is crucial for the service-oriented remote-sensing data and the related processing functions to develop a more effective method of semantic sharing and interoperability in order to satisfy actual applications in a high-efficiency, cooperative and precise way.This dissertation mainly addresses remote sensing information processing services to develop sharing and interoperability at the semantic level.Based on the current research situation and inherent characteristic of service-oriented remote-sensing data processing, this dissertation employs a series of theoretical and practical methods, such as ontology theory, semantic web technology, linguistic theory, and formal concept analysis, to investigate semantic description of remote sensing information processing services. The primary contents are as follows:(1) Semantic modeling approach for remote sensing information. As the object dealt with by service, remote sensing information is indispensable to semantic description of remote sensing processing services. Only fully modeling the semantic description of remote sensing information can characterize the semantics of remote sensing information processing services completely. This dissertation clarifies why and how to model the semantics of remote sensing information. Through introducing the ontology theory, a novel ontology-based semantic modeling approach of remote sensing information is proposed. Subsequently, this approach is elaborated in detail from these aspects including the definition, component elements, construction method, and formalized representation of remote sensing information ontology, which provides the underlying basis for the following semantic representation of remote sensing information processing services.(2) Classification ontology and the corresponding construction method for remote sensing information processing services. By analyzing the existing taxonomic hierarchies for geo-spatial information services as well as the traditional information classification methods, this dissertation introduces the conception of ontology-based classification of remote sensing information processing services. Subsequently, in view of the frame semantics theory and componential analysis theory in linguistics, as well as the domain features and linguistic characteristic of remote sensing information processing services, an approach is developed to build remote sensing information processing services classification ontology based on event frame and formal concept analysis, which is able to provide supports to the semantic representation of remote sensing information processing services.(3) Ontology-based semantic description model for remote sensing information processing services. By comparing the existing description methods or languages, and analyzing the fundamental contents and principles of semantic description regarding remote sensing information processing services, this dissertation puts forward an ontology-based semantic description model, which overcomes the shortcomings of OWL-S and lays the semantic foundations for automatic discovery, composition and invocation of remote sensing information processing services.(4) Registration problem of ontology-based semantic description for remote sensing information processing services. Addressing the intrinsic shortage that the existing catalogue services for web don’t support the semantic description for remote sensing information processing services, a scheme of semantic extension to the existing registry information model is developed, and accordingly a prototype design concerning semantic registration center of remote sensing information processing services is carried out in this dissertation, which can offer registration supports for discovery and composition arising in the semantics-based remote sensing information processing services.As described above, this dissertation develops a set of theoretical framework and practical method to address ontology-based semantic description of remote sensing information processing services. The main innovative points are summarized as follows:(1) A semantic modeling approach for remote sensing information is proposed and a formal definition of remote sensing information ontology is given in the dissertation. Depending on the essential features of remote sensing information, a nine-step construction method is presented to set up remote sensing information ontology, which is crucial for semantic representation of remote sensing information processing services.(2) A novel approach is put forward to construct remote sensing information processing services classification ontology based on event frame and formal concept analysis. In combination with the frame semantics and componential analysis, this approach can extract semantic features from concepts of remote sensing information processing services, and then exploit the formal concept analysis to establish classification ontology for remote sensing information processing services.(3) A novel ontology-based semantic description model for remote sensing information processing service is brought forward. On the one hand, this model can distinguish abstract functional description, concrete service instance description and invocation description, such that remote sensing information processing services can be characterized from different description levels, which will provide necessary supports for automatic discovery, composition and invocation of services. On the other hand, this model represents the semantics of remote sensing information processing services through introducing both remote sensing information ontology and remote sensing information processing services classification ontology. Furthermore, it also adds the semantics about precision and Qos of remote sensing information processing services.(4) A semantic registration information model is proposed, which is combined with ontology-based semantic description model to give a significant semantic extension to ebRIM. Likewise, the model is capable of supporting the registration to semantic descriptions of remote sensing information services, which paves the way for discovery, composition and invocation of semantics-based remote sensing information processing services.To be able to apply the results of this research into distributed platforms of geospatial information services, the following issues should be studied and addressed further:(1) Method for ontology integration. Different geospatial information communities usually build their own remote sensing information ontologies and processing service operation ontologies, leading to semantic heterogeneity between different ontologies, thus need to research methods for ontology integration to solve the problem.(2) Assessment method for the proposed ontology-based semantic description model of remote sensing information processing services, and improve the proposed model based on the assessment results.(3) Method for semantic-based discovery and composition of remote sensing information processing services, thus to support automatic and semi-automatic aggregation of remote sensing information processing services distributed on the web to complete complex application purpose. |
PDF Full Text Request