مقايسه رويكردهاي استخراج روابط توپولوژي سه بعدي در نسل هاي مختلف سيستم اطلاعات مكاني

Comparison of approaches to extract the three-dimensional topological relations in different generations of GIS

نسل جديدي از سيستم اطلاعات مكاني (GIS)، نسل GIS فراگستر است كه در اين نسل قابليت سرويس دهي به هر كاربر در هر مكان و زماني داده مي شود. با پيشرفت فناوري هاي جديد در اخذ اطلاعات سه بعدي از محيط تحولي در نسل هاي قديم سيستم اطلاعات مكاني ايجاد شده است. يكي از موارد حائز اهميت در نسل هاي مختلف سيستم اطلاعات مكاني، استخراج روابط مكاني است. استخراج روابط توپولوژي سه بعدي كه بيشتر در سيستم اطلاعات مكاني موردتوجه قرارگرفته است، مبتني بر مدل داده هاي استفاده شده در نسل هاي مختلف GIS صورت مي گيرد. مدل داده هاي مورداستفاده در نسل هاي قديم GIS داراي محدوديت هايي همچون عدم تعامل پذيري بالا بين كاربر و محيط، عدم نمايش سه بعدي اشياء به دست آمده از محيط و هم چنين ناسازگاري اشياء با پيش فرض هاي استفاده شده هستند، درحالي كه در نسل GIS فراگستر با استفاده از حسگرها و زيرساخت هاي مختلف در محيط، محدوديت هاي بيان شده براي سرويس دهي برطرف مي گردد. هدف اين مقاله مقايسه ي رويكردهاي مورداستفاده در نسل هاي مختلف سيستم اطلاعات مكاني براي استخراج روابط توپولوژي مبتني بر مدل داده هاي ارائه شده در اين نسل ها است. در اين مقاله استخراج روابط توپولوژي در نسل هاي سنتي سيستم اطلاعات مكاني، با استفاده از روش مدل سازي شئ گرا و شبيه سازي اشياء هرم و كره صورت مي گيرد. اين در حالي است كه در نسل GIS فراگستر از داده هاي به دست آمده توسط كينكت به عنوان يك حسگر انتخابي مبتني بر مدل داده ي ارائه شده، روابط استخراج مي گردد. نتايج به دست آمده از مقايسه ي رويكردهاي استخراج روابط توپولوژي سه بعدي در اين دو نسل، بيانگر افزايش تعامل كاربر با محيط پيرامون مبتني بر ابزارهاي موجود در نسل هاي فراگستر براي استخراج روابط توپولوژي است.

Today، ubiquitous Geospatial Information Systems (GIS) has been growing to become the recent generation of GIS. In this generation of GIS، servicing capabilities to any user، at any time، in any location، using any device، and in any condition are provided. Advances in Information Technology (IT) industry have led to the advent of low-cost three dimensional data acquisition technologies (e.g. Microsoft Kinect) that changes the approaches of current generation of GIS. One of the important matter in different generation of GIS is 3D spatial relations extraction. 3D topological extraction used more in GIS، is done based on the data models in different generation of GIS. There are some deficiencies and limitations in current data models such as lack of high interoperability between the user and the environment، lack of visualizing 3D objects relationships with each other and inconsistency between 3D objects and their primitives، while in ubiquitous generation of GIS by using different sensors and infrastructures in the environment these deficiencies are improved. In this article for extracting 3D topological relations in current GIS data models، the methodology of object-oriented procedure and simulation of different objects like sphere and pyramid are done. By the way، for the ubiquitous generation of GIS، by using smart Kinect sensor، extraction of 3D topological relations is done for a building prototype. Information Technology (IT) has faced a change in its earliest style of merely making an interaction between the users and the environment within a computer. Nowadays، a new era has arrived at which people can communicate with their environment through embedded sensors on the basis of ubiquitous computing. With the advent of novel technologies of three dimensional information acquisition from the environment، a revolution in the ubiquitous computing has occurred which remarkably broadens its capabilities. One of these technologies is the use of Microsoft Kinect smart sensor to acquire the depth information in any location and at any time. One of the requirements of Ubiquitous GIS is the use of these sensors capabilities in the extraction of spatial relations that empower us with some spatial analysis. The purpose of this article is the comparison of approaches to extract 3D topological relations based on the data models used in different generations of GIS. The use of an appropriate spatial data model for ubiquitous computing makes the extraction of spatial relations at any time and in any location possible as a service for the users. Topological relations are considered an important property of spatial objects، and are fundamental to spatial analyses and query-based processing. As the information collected by Kinect is in the form of 3D point clouds، the extraction of three dimensional topological relations is feasible. The spatial analyses describe the topological relations based on some predefined models which are generated at the same time with the process of modeling the objects. In this study، based on a ubiquitous data model and using Kinect as a sensor utilized in ubiquitous computing to enhance the interactivity between users and the environment and to acquire 3D information and to extract topological relations. The proposed data model، termed Ubiquitous GIS Data Model (UGDM)، is able to support various spatial relations in a metric or non-metric manner. UGDM well supports various spatial relations، including directional، distance-based، projection، and topological relations. The extraction of eight topological relations via different sensors is among the capabilities of this data model. The results of this article for provided data indicates the difference between the generations of GIS to extract topological relations، and more interaction between user and environment in ubiquitous generation of GIS by different sensors. The extraction of spatial relations between objects provides us with substantial information about their surrounding environment. Directional، distance based and also topological relations are some instances of spatial relations. The used methodology in this paper according to the ubiquitous GIS generation starts with the acquisition of information in the form of point clouds. Then topological information like algebraic topology and point set topology are extracted. Providing a ubiquitous data model for extracting 3D topological relations based on different sensors and infrastructure enhances the interaction between the user and the environment. This capability in the presented data model shows the lack of interoperability in the current data models. Providing a language for presented data model in this article is the paper future goal.