The Research Of Reasoning Of Position Information And Modelling For Three Dimensional Directions Relation Based OPRA

Spatial reasoning is an important part of artificial intelligence, geographic informationsystems, robotics, computational vision, image retrieval, natural language processing andother fields. The present study of spatial reasoning mainly focus on representation andinference for a single spatial relationships of static spatial objects and two-dimensionalspatial objects. In reality, however, most of the spatial information is dynamic and sometimeswe shoud consided many spactial relations.With the research of spatial reasoning applied inlife,we shoud conside the directions relation in three dimensional space. OPRA theory andapplication research has matured, but only for OPRA we consided one spatial relation and theobject in the two-dimensional space. In this paper,we give the the research of reasoning ofposition information and modelling for three dimensional directions relation based OPRA.Theposition reason is only for static object and is small range,the proposed position reason in thispaper can be applied in dynamic scenario and a large range.Space object is located in thethree-dimensional space, not only in the plane of orientation relationship,we also need tostudy on high position relation, in this paper, based on the two-dimensional OPRA, we studiesthe relationship between the three dimensional relative orientation model, and this modelformalized representation car navigation problems.Details are as follows:(1) This paper sums up and analyzes on current state of the relative direction relations,the position model and reason of combined with the direction and distance relationship andthe relative direction relation model in three-dimensional space.At the same time, Given thepurpose and significance of this paper.(2) This paper introduced the relevant background knowledge. First, a brief introductionto several common models of the relative orientation relationship between point objects;followed by a detailed description of the oriented points of relational algebra, relative motionorientation relationship model, the qualitative distance relationship; Finally, give the basicconcepts of relational operators.(3) First, the direction of movement about moving point is the internal direction of theoriented point, we combined the model of OPRA with the relative motion orientationrelationship model.Through the appropriate direction angle value,we analysis the relationshipbetween the model of OPRA and the relative motion orientation relationship model.Then,wegive the position relationship model which combined the relative direction relations, therelative motion orientation relationship, the distance relationship.At the same time,we give the range of the angle of the composition of distance relationship by the OPRA.At the last,wegive the algorithm of the position relationship reasoning.(4) In order to express the relative direction relation in the three-dimensional space, Therelative direction in which is a combination of xoy plane and z-axis direction,at the sametime,this paper proposed three-dimensional model3DOPRAmby extending the modelOPRAmfor two-dimensional relative direction relation and respectively defining the new tworelative angles.Moreover, the formal definition of3DOPRAmmulti-granularity and its inverserelationship are given.we also give the conceptual neighborhood of3DOPRAmwhich appliedin practice. Finally, the conceptual neighborhood based on the action of3DOPRAmmodel isapplied to describe the car navigation problems.(5) Design and implement the the algorithm of the position relationship reasoning and thedemonstration system of the model of position information.The model of position information about moving object we proposed can be applied bythe moving object,at the same time,the model can give a more accurate spatial informationwhich makes the spatial scene description more accurate. Compared with the existingtwo-dimensional model,3DTPCC is more expressive and can describe more directions inthree-dimensional space, not only relations in two-dimensional space. The work of this paperhas the theoretical significance and application value of the spatial reasoning combining withvarious spatial relations,which is applied in spatio-temporal GIS, robot navigation, and so on.