Presentation is loading. Please wait.

Presentation is loading. Please wait.

Peter BogaertGeos 2005 A Qualitative Trajectory Calculus and the Composition of its Relations Authors: Nico Van de Weghe, Bart Kuijpers, Peter Bogaert.

Similar presentations


Presentation on theme: "Peter BogaertGeos 2005 A Qualitative Trajectory Calculus and the Composition of its Relations Authors: Nico Van de Weghe, Bart Kuijpers, Peter Bogaert."— Presentation transcript:

1 Peter BogaertGeos 2005 A Qualitative Trajectory Calculus and the Composition of its Relations Authors: Nico Van de Weghe, Bart Kuijpers, Peter Bogaert and Philippe De Maeyer Department of Geography, Ghent University, Belgium {nico.vandeweghe, peter.bogaert, philippe.demaeyer}@ugent.be Theoretical Computer Science, Hasselt University, Belgium bart.kuijpers@uhasselt.be

2 Peter BogaertGeos 2005 Overview  Problem statement  The Qualitative Trajectory Calculus – Double Cross (QTC C )  Composition Table  Composition Rules Table  Concluding Remarks

3 Peter BogaertGeos 2005 Problem Statement Qualitative Reasoning: Region Connection Calculus (RCC) Database: 9-Intersection Model Randell, D., Cui, Z, and Cohn, A.G., 1992. A Spatial Logic Based on Regions and Connection, In: Nebel, B., Swartout, W., and Rich, C. (Eds.), Proc. of the 3rd Int. Conf. on Knowledge Representation and Reasoning (KR), Morgan Kaufmann, San Mateo, USA, 165 ‑ 176. Egenhofer, M. and Franzosa, R., 1991. Point ‑ Set Topological Spatial Relations, International Journal of Geographical Information Systems, 5 (2), 161 ‑ 174. a lot of work has been done in stating the dyadic topological relations between two polygons Two approaches

4 Peter BogaertGeos 2005 Problem Statement DC(k,l)EC(k,l)PO(k,l) TPP(k,l)NTPP(k,l) TPPI(k,l)NTPPI(k,l) EQ(k,l) 8 possible topologic relations Constraints upon the ways these relations change

5 Peter BogaertGeos 2005 Problem Statement Developing a calculus for representing and reasoning about movements of objects in a qualitative framework. continuously moving objects: often only “disconnected from” “How do we handle changes in movement between moving objects, if there is no change in their topological relationship?” DC(k,l)

6 Peter BogaertGeos 2005 QTC (Van de Weghe, N. (2004) Representing and Reasoning about Moving Objects: A Qualitative Approach, PhD thesis, Belgium, Ghent University, Faculty of Sciences, Department of Geography, 268 pp.) QTC B 1D 2D Distance, speed QTC C 2D Double Cross Concept A calculus for representing and reasoning about movements of two disconnected point-like objects in a qualitative framework.

7 Peter BogaertGeos 2005 QTC – Simplification

8 Peter BogaertGeos 2005  Freksa, Ch., 1992. Using Orientation Information for Qualitative Spatial reasoning, In: Frank, A.U., Campari, I., and Formentini, U. (Eds.), Proc. of the Int. Conf. on Theories and Methods of Spatio ‑ Temporal Reasoning in Geographic Space, Pisa, Italy, Lecture Notes in Computer Science, Springer ‑ Verlag, (639), 162 ‑ 178. Double-Cross Concept

9 Peter BogaertGeos 2005 QTC C Double-Cross Calculus (Freksa) QTC – DOUBLE-CROSS (QTC C ) l k

10 Peter BogaertGeos 2005 Double-Cross Calculus (Freksa) QTC C l k l k QTC – DOUBLE-CROSS (QTC C )

11 Peter BogaertGeos 2005 Double-Cross Calculus (Freksa) QTC C l k l k l k QTC – DOUBLE-CROSS (QTC C )

12 Peter BogaertGeos 2005 QTC – DOUBLE-CROSS (QTC C ) l k

13 Peter BogaertGeos 2005 0 – + – QTC – DOUBLE-CROSS (QTC C ) l k

14 Peter BogaertGeos 2005 – + –– 0 QTC – DOUBLE-CROSS (QTC C ) l k

15 Peter BogaertGeos 2005 0 – ––– + QTC – DOUBLE-CROSS (QTC C ) k l

16 Peter BogaertGeos 2005 0 – + –––– – + QTC – DOUBLE-CROSS (QTC C ) l k

17 Peter BogaertGeos 2005 Qualitative Trajectory Calculus (QTC) QTC B2D QTC C

18 Peter BogaertGeos 2005 Qualitative Trajectory Calculus (QTC) QTC B2D QTC C

19 Peter BogaertGeos 2005 Qualitative Trajectory Calculus (QTC) QTC B2D QTC C

20 Peter BogaertGeos 2005 R 1 (k,l)  R 2 (l,m) = R 3 (k,m)  originate from temporal reasoning (Allen 1983)  encodes all possible compositions of relations  simple table look ‑ up  useful from a computational point of view Composition Table  Idea: to compose a finite set of new facts and rules from existing ones

21 Peter BogaertGeos 2005 Composition Table

22 Peter BogaertGeos 2005 Composition Table  QTC C defines 81 different relations  Difficult visual presentation  A composition table would have 81*81 (6561) entries

23 Peter BogaertGeos 2005  Useful for visual presentation  Instrument to help generate the composition table Composition Rules Table  Reduction of the number of entrances

24 Peter BogaertGeos 2005 Composition Rules Table  Based on 2 rules  Which rotation do we need, such that l of R2 matches l of R1?  How is k moving with respect to l in R1, and how is m moving with respect to l in R2?

25 Peter BogaertGeos 2005  Which rotation do we need, such that l of R2 matches l of R1? Composition Rules Table (k,l)(– + + –) C (l,m)(– – – +) C k l l m (k,m)

26 Peter BogaertGeos 2005 (case 1 of n) (k,l)(– + + –) C (l,m)(– – – +) C k l l m l (k,m) k m Composition Rules Table

27 Peter BogaertGeos 2005 (k,l)(– + + –) C (l,m)(– – – +) C (case 1 of n) (k,m) k m k l l m Composition Rules Table

28 Peter BogaertGeos 2005 m k – (k,m)(– ) C (k,l)(– + + –) C (l,m)(– – – +) C (case 1 of n) (k,m) k l l m Composition Rules Table

29 Peter BogaertGeos 2005 k l l m m k + (k,m)(– + ) C (k,m)(– ) C (k,l)(– + + –) C (l,m)(– – – +) C (case 1 of n) (k,m) Composition Rules Table

30 Peter BogaertGeos 2005 m k – (k,m)(– + ) C (k,m)(– ) C (k,m)(– + – ) C (k,l)(– + + –) C (l,m)(– – – +) C (case 1 of n) (k,m) k l l m Composition Rules Table

31 Peter BogaertGeos 2005 k l l m m k + (k,m)(– + ) C (k,m)(– ) C (k,m)(– + – ) C (k,m)(– + – + ) C (k,l)(– + + –) C (l,m)(– – – +) C (case 1 of n) (k,m) Composition Rules Table

32 Peter BogaertGeos 2005 k l l m m k l (k,l)(– + + –) C (l,m)(– – – +) C (case 2 of n) (k,m) Composition Rules Table

33 Peter BogaertGeos 2005 k l l m k (k,l)(– + + –) C (l,m)(– – – +) C (case 2 of n) (k,m) m Composition Rules Table

34 Peter BogaertGeos 2005 k l l m k (k,l)(– + + –) C (l,m)(– – – +) C (case 2 of n) (k,m) m (k,m)(– – + +) C Composition Rules Table

35 Peter BogaertGeos 2005 (k,l)(– + + –) C (l,m)(– – – +) C > 180° k l m l (k,m) l (n of n cases) Composition Rules Table

36 Peter BogaertGeos 2005 (k,l)(– + + –) C (l,m)(– – – +) C > 180° < 360° k l m l l k l m (n of n cases) (k,m) Composition Rules Table

37 Peter BogaertGeos 2005 > 180° and < 360° (n of n cases)(k,l)(– + + –) C (l,m)(– – – +) C (k,m) Composition Rules Table

38 Peter BogaertGeos 2005 Which rotation do we need, such that l of R 2 matches l of R 1 > 180° and < 360° (n of n cases)(k,l)(– + + –) C (l,m)(– – – +) C (k,m) Composition Rules Table

39 Peter BogaertGeos 2005 How is k moving with respect to l in R 1, and how is m moving with respect to l in R 2 (n of n cases)(k,l)(– + + –) C (l,m)(– – – +) C (k,m) Composition Rules Table

40 Peter BogaertGeos 2005 How is k moving with respect to l in R 1, and how is m moving with respect to l in R 2 (n of n cases)(k,l)(– + + –) C (l,m)(– – – +) C (k,m) Composition Rules Table

41 Peter BogaertGeos 2005 Composition Rules Table

42 Peter BogaertGeos 2005  independent  goal: movement of object k with respect to object n (R 3 )  incomplete data  incomplete answer  two scientific teams 'Puzzling the Past' (k,n)?

43 Peter BogaertGeos 2005 (k,l)(– + + –) C (l,n)(– – – +) C Team I Team II (k,m)(– + – +) C (m,n)(– + – +) C k l l m m n n k (k,n)(k,n) (k,n)(k,n) 'Puzzling the Past'

44 Peter BogaertGeos 2005 Team IIITeam IV Team I Team II   Team III Team IV (k,n)(k,n) 'Puzzling the Past'

45 Peter BogaertGeos 2005 Team V (k,n)(– + – +) C Team IIITeam IV 'Puzzling the Past'

46 Peter BogaertGeos 2005 Complex researches (huge number of anchor points, teams, measurements per team, updates, etc.) Cuncluding Remarks  IMPLEMENTATION OF QTC C The composition-rules table forms a basis for reasoning about incomplete spatio-temporal knowledge in information systems. Can be used in a variety of research domains: geomorphology, geology, archaeology, and biology.

47 Peter BogaertGeos 2005 A Qualitative Trajectory Calculus and the Composition of its Relations Authors: Nico Van de Weghe, Bart Kuijpers, Peter Bogaert and Philippe De Maeyer Department of Geography, Ghent University, Belgium {nico.vandeweghe, peter.bogaert, philippe.demaeyer}@ugent.be Theoretical Computer Science, Hasselt University, Belgium bart.kuijpers@uhasselt.be


Download ppt "Peter BogaertGeos 2005 A Qualitative Trajectory Calculus and the Composition of its Relations Authors: Nico Van de Weghe, Bart Kuijpers, Peter Bogaert."

Similar presentations


Ads by Google