Web-Books
im Austria-Forum
Austria-Forum
Web-Books
Informatik
Joint Austrian Computer Vision and Robotics Workshop 2020
Seite - 138 -
  • Benutzer
  • Version
    • Vollversion
    • Textversion
  • Sprache
    • Deutsch
    • English - Englisch

Seite - 138 - in Joint Austrian Computer Vision and Robotics Workshop 2020

Bild der Seite - 138 -

Bild der Seite - 138 - in Joint Austrian Computer Vision and Robotics Workshop 2020

Text der Seite - 138 -

clidean topology and f will describe a continuous image or3D-scan. Definition2.1. Apath isacontinuousfunctionfrom the real interval [a,b] (witha<b) into a topological spaceΩ. Definition 2.2. Two points x 6= y in a topological spaceΩarecalledpath-connected ifandonlyif there exists a path γ : [a,b]→ Ω with γ(a) = x and γ(b) =y. Definition 2.3. The set of all points which are path- connected to a pointx∈Ω is the connected compo- nentofx: [x] :={y∈Ω|x ispath-connected toy} Any subset of Ω which can be written in above way (forasuitablechoiceofx) iscalledaconnectedcom- ponent. Definition2.4. Apathγ : [a,b]→Ω iscalledmono- tonic if andonly if thewholepath isascendingor the wholepathisdescending,meaningthefirstorsecond formula belowhas tohold, respectively: ∀s,t∈ [a,b] :s<t⇒f(γ(s))≤f(γ(t)) ∀s,t∈ [a,b] :s<t⇒f(γ(s))≥f(γ(t)) Definition 2.5. LetR ⊂ Ω. R is called slope re- gionormonotonicallyconnected if andonly if forall x,y∈R there exists a monotonic pathγ : [a,b]→ Rwithγ(a) =xandγ(b) =y. Definition 2.6. A family of sets{Ai⊂Ω | i∈ I} is called a sloperegiondecomposition if andonly if: • Ai is a slope region for all i∈ I • ∀i,j∈ I : i 6= j⇒Ai∩Aj=∅. • ⋃i∈IAi= Ω Definition 2.7. Consider two slope region de- compositions A = {Ai⊂Ω | i∈ I} and B = {Bj⊂Ω | j∈J}. We callA coarser thanB, al- ternativelyBfiner thanA, in SymbolsA B if and only if ∀j∈J ∃i∈ I :Bj⊂Ai. Theorem2.8. isapartialorder, i.e. fulfills reflex- ivity, antisymmetryand transitivity. Proof: Straight forward. Antisymmetry follows fromthedecompositionproperty. Definition 2.9. A slope region decompositionA is calledmaximallycoarseorsimplycoarse ifandonly if there is no other coarser slope region decomposi- tion. We can apply Zorn’s lemma [6] to the partial or- der , which yields the existence of maximal ele- ments. For this we need to show that chains have upperbounds. Theorem2.10. Foranyascendingchainof slopere- gion decompositions (Ai)i∈I, that is t≥ s⇒At As, there is a slope region decompositionA∞ satis- fying∀i∈ I :A∞ Ai. Proof: Weconsider theequivalencerelation”con- nected inAi” for twopointsx,y∈Ω: x∼i y :⇔∃A∈Ai :x∈A∧y∈A Theequivalence relation is a subsetofΩ2, andAt As implies∼t⊃∼s. This suggests the use of∼∞:=⋃ i∈I ∼i to get an upper bound. Indeed the equiv- alence classes of∼∞ yield a partitionA∞ of Ω, which is coarser than anyAi. But do they form a slope region decomposition? Yes: For any two fixed pointsx,y to be∼∞-connected, they need to be∼i- connected for some i ∈ I. So there is a mono- tonic path linkingx and y inA= [x]∼i ⊂ [x]∼∞, by which they are monotonically connected inA∞. ThereforA∞ is a slope regiondecomposition. HenceeverysetΩhasacoarsedecomposition. Theorem 2.11. LetA⊂Ω be a path-connected set. A is a slope region if and only if all levelsets off in Aarepath-connected, i.e. ∀c∈R :f−1({c})∩A ispath-connected. Proof: ”⇒”viacontraposition: Supposethereexistsac∈RwithL :=f−1(c)∩A not path-connected. We decomposeL in its compo- nents and pick x and y from different components. Since f(x) = f(y) = c a monotonic path between xandywouldhave to lie completely inL.However, since x and y are from different components, they cannotbeconnectedbyapathinLandthereforecan- not be connected with a monotonic path. Therefore, A isnot a slope region. ”⇐”via ironingout anarbitrarypath: Givenx,y∈Awe have to find a monotonic path γ. Without loss of generality supposef(x)≥ f(y). SinceA is path-connected, there exists an (not nec- essarily monotonic) path γ0 : [a,b]→A fromx to 138
zurück zum  Buch Joint Austrian Computer Vision and Robotics Workshop 2020"
Joint Austrian Computer Vision and Robotics Workshop 2020
Titel
Joint Austrian Computer Vision and Robotics Workshop 2020
Herausgeber
Graz University of Technology
Ort
Graz
Datum
2020
Sprache
englisch
Lizenz
CC BY 4.0
ISBN
978-3-85125-752-6
Abmessungen
21.0 x 29.7 cm
Seiten
188
Kategorien
Informatik
Technik
Web-Books
Bibliothek
Datenschutz
Impressum
Austria-Forum
Austria-Forum
Web-Books
Joint Austrian Computer Vision and Robotics Workshop 2020