Showing papers by "Jagannath Aryal published in 2012"

Association of Spatial Object Recognition and GEOBIA: an Experimental Approach

Abstract: Spatial object recognition is an important facet of spatial information system. Psychological and mathematical basis are two important factors in forming spatial object. In recent years, Geographic Object Based Image Analysis (GEOBIA) is extensively used in extracting the information from remotely sensed images. The extracted information is being used by incorporating the experts’ knowledge. With these developments, in this paper, we conducted an experiment for the human subjects in an image of an area of natural importance. The image along with eight different sets of segmented images of the same area were provided to the human subjects, such that they can improve their understanding using their perception, cognition and decision for the Geographic object recognition. The results of two stages were collated and a relation to mathematical and psychological basis is made. The results showed that there exists different ways to comprehend the geographical objects according to background experience of human subjects, despite the use of similar automatic spatial partitions at different scales. But, also that there exists a certain association of pertinent scale to the human perception in recognising the spatial objects.

Solving the optimal location problem in forest fire control with fuzzy data points

Abstract: In this paper, we present a methodology to solve location problems when the data used is inherently fuzzy. This method, from data clusterized with the fuzzy c��means algorithm, calculates bi-dimensional fuzzy numbers from the clusters which are used to calculate a fuzzy solution. We apply the methodology, with different objective functions, to a particularly apt data set of forest fire breakouts in the Bouches du Rhone region of southern France, gathered from 1981 to 2009. The robustness of the method is then evaluated with a Monte Carlo simulation in which the number of clusters change. The solution provided with this fuzzy method provides leeway to planners, which can see how the membership function of the fuzzy solution can be used as a measurement of "appropriateness" of the final location.

3D modelling and analysis: ISO standard tools for air traffic

Abstract: The Geographic Information Systems (GIS) applied to aviation use mostly modelling and analysis in 2D. Nevertheless, a tendency to represent and analyse in three-dimensions begins to emerge. It requires new descriptions and new operative tools for 3D objects in GIS. As an association of users and software producers, the Open Geospatial Consortium (OGC) has defined an ISO standard to describe two-dimensional and three-dimensional geometric objects. However, this standard does not permit a description of common Computer Aided Design (CAD) objects, a vital task remains for the modelling of 3D data and their uses, in the context of analysis (spatial query) or with the proposals of new primitives. In this research, our goal is to build a computing library fully compatible with ISO standard especially allowing characterising any gaps or parts requiring further development. In this paper, we present a solution validated by a use-case for modelling and analysis of aerial traffic in 3D on an ellipsoid of revolution that follows the ISO standard. In order to demonstrate whether new geometric objects that we propose are effective, in this simple approach, we have established a complete processing chain.

Scalar data analysis for assessing biodiversity of various geographical territories using GEOBIA

Abstract: Geographic Object Based Image Analysis (GEOBIA) is us ed in the advancement of spatial object recognition fr m wider community including landscape analysts. Due to the spatial co mponent inherent in the landscape, the relationship of landscape phenomena to remote sensing and object recognition is well recog nised. The landscape phenomena exist and interact i n multiple scales. The interaction in multiple scales occurs within the sc ale and across the scales. To address the issue of this interaction, we developed a scalar data analysis (ScDA) framework in multi-scal e environment from remotely sensed data of diverse geographical territories (New Zealand, Nepal and France) by extracting the m eaningful image objects, analysing such image objec ts and relating these image objects to landscape objects. ScDA was applied for the indices such as the Normalised Difference Veget ation Index (NDVI), the Grey Level Co-occurrence Matrix (GLCM), shape index, area, density and asymmetry for image objects. Thes e indices and the developed framework were tested for pertinent scale (th most appropriate scale for analysis) issues u sing a statistical measure of association – The Relative Interquartile Range (RIQR) a nd an algorithmic approach. The test result showed that the most appropriate scale to analyse -pertinent scalecan be achieved and is dependent primarily on analysis and interpre tation of the objects which are governed by perception, recognition as well as obje ctiv of the interpreter / analyst including hetero geneity / homogeneity of the landscape. This methodology showed that pertinent s cale issue is relevant for the study of biodiversit y monitoring and associated landscape phenomena. * Corresponding author.