Merken

# The way to go with WPS

#### Automatisierte Medienanalyse

## Diese automatischen Videoanalysen setzt das TIB|AV-Portal ein:

**Szenenerkennung**—

**Shot Boundary Detection**segmentiert das Video anhand von Bildmerkmalen. Ein daraus erzeugtes visuelles Inhaltsverzeichnis gibt einen schnellen Überblick über den Inhalt des Videos und bietet einen zielgenauen Zugriff.

**Texterkennung**–

**Intelligent Character Recognition**erfasst, indexiert und macht geschriebene Sprache (zum Beispiel Text auf Folien) durchsuchbar.

**Spracherkennung**–

**Speech to Text**notiert die gesprochene Sprache im Video in Form eines Transkripts, das durchsuchbar ist.

**Bilderkennung**–

**Visual Concept Detection**indexiert das Bewegtbild mit fachspezifischen und fächerübergreifenden visuellen Konzepten (zum Beispiel Landschaft, Fassadendetail, technische Zeichnung, Computeranimation oder Vorlesung).

**Verschlagwortung**–

**Named Entity Recognition**beschreibt die einzelnen Videosegmente mit semantisch verknüpften Sachbegriffen. Synonyme oder Unterbegriffe von eingegebenen Suchbegriffen können dadurch automatisch mitgesucht werden, was die Treffermenge erweitert.

Erkannte Entitäten

Sprachtranskript

00:00

was

00:05

so WPS the Web Processing Service of course is the OGC standard for computational oriented of Web services has much in common with particles like the WMS and WFS what we'd support for of a synchronous request suitable for long-running processes and also wrote support for a nested requests so my presentation is mainly about a particular application we use we we as to serve the results and also an example of using nested requests would be this and then the application is it's a path planning

00:56

problem for on the way to go In other words to find the best route between 2 points were best may be the quickest the shortest the most economical with the least chance of accidents for example and usually performed on a network like this road network that is pictures here with nodes edges that connects them and weights on each edge has however but we would like

01:28

to go places like this or

01:31

this and basically anywhere you can go on foot or with the ground legal so so we want to build the

01:43

path planning application for training and initially to cover all of my and my country Norway which is about 2 thousand kilometers long and find the safest route and not restrict motion to roads of paths and estimate travel time and perhaps other things like exposure visibility and I'm going to use this small area north of Mosul 5 by 5 kilometers should examples on how we build the graph to run for and so our

02:21

perspective is that we need to do a situation-dependent path planning in a large graph situation-dependent mean that means that the weights may change of according to the time of the situation it might be a weather dependent for example seasons dependence on I also we have to relate to us service-oriented information infrastructure and that's where wp is constant I we have very detail land cover data on the type of vegetation support and we're sort of expecting in the coming years to have access to a very high resolution elevation data on a national scale so we're trying to tailor the application to have the potential that's inherent in that kind of data and also we are working on simulations of ground physics so that we can also predict things like temperature profiles and around and load carrying capacity of the ground so that's our starting point so here on

03:41

going to put the most weight on the on the graph generation show some routine examples and say a little bit about the years in the implementation

03:54

so I have to run out of which create a graph that's covers all of all of the land there are various possibilities you can have a regular graph like this perhaps with diagonals as well a very coarse grid shown here but it tends to produce a great number of nodes so you can have a random graph where the nodes are distributed more directly and there's the graph visibility graph where the nodes are usually placed close to obstacles and you assume that you can move freely on the edges between nodes connecting those nodes on the same so the Voronoi graph which is also used in robotics for example typically then the centroid in the graph in each cell is uh at an obstacle and you want to move along the edges over the cells to be as far as possible away from the obstacles however the for persons moving into rain you often find that you mean you pass obstacles as close as possible and finally both there are more but navigation measures which I believe are popular in gaming the game and straight it's also made from polygons where you assume that you can move freely within each polygon and the edges connect adjacent columns so our approach is to use

05:42

and we the random graphs but we somewhat judiciously distributed nodes in other words the nodes positions are sampled from some distribution that reflects the terrain properties so the more nodes were more nodes are needed to put so that's a graph types and so we also have some different

06:09

data types to produce the grade so elevation of course I went from elevation data we also derive attributes that describe the during more succinctly like this is what I

06:27

call actual variance which basically tells you how much your surface normal lecture jumps around in the neighborhood and other attributed and you do right as well so then there's the land cover

06:43

data I just showed here as polygons to give an impression of the detail this is a data set that's used a lot in in forestry so it describes for example the potential yield all the forest but it has the the data about soil type agitation time or whether it's a naked roads on etc. so then there is a road networks of course and path trail networks and for the latter we use that for the time being reuse OpenStreetMap data and then there's the uh that's basically what we are using for the moment we also plant implement the latter categories especially lighter ground physics and weather forecasts this is

07:39

another view of a photograph of the area that cover problems so this 3 lakes in the middle of the forest all this is

07:52

an example of a later late ly the datasets we haven't got so much of that yet so we just tested it's very exciting to work with gives a whole different Our characterization of round morphology and the vegetation

08:12

so of guiding principles then is to go use create a random graph with no density depending on during attributes we use a trying relation to form the edges but we want to our have that but have the ability to have a great number of nodes the woods arbitrary of spatial resolution and also on an arbitrary geographical scale so that's why we introduce a hierarchical representation of its of having our graph in several the levels of layers where each layer of the unknown in 1 layer is formed from a cluster of nodes in more detail later so what what

09:09

to to generate the graph the 1st lay out the uh the roads data what we serve all vertices which I used what for computing the weights but we convert some of the vertices tune to in graph nodes as well so we have extra nodes to connect to the rest of the graph we do the same with the path constant and then we distribute altering notes this

09:39

so in my move your friend here course resolution but it's search for an illustration purposes so we

09:49

generate the graph in this example we completely avoid lakes and marshes so that in the winter time of course we want to include that and just to reflect that the surface in the edge weights

10:03

are so working with gas very convenient to use represented as a sparse matrix or each of each node is a role and a column in there large matrix with many zeros and non-zero elements represent connected nodes and the diagonal reflected in the case of it's a directed graph the hierarchical

10:32

representation simple idea of merging of nearby nodes nearby in the sense of our distance measures which which is travel time in this application so that we can I start happening very on the course level and pretty 3 a shortest path algorithm on the solution at the of the union of all the nodes in the shortest path solution of the course level perhaps including neighboring neighbors of neighboring clusters so it's a it's a way of bridging been reducing the number of nodes that enters into the computation on how 1 of the reasons we do this but if you recall we we want to do but a dynamic company up the computation of the of the weights and the weights are a situation-dependent so the we cannot of update the whole a gigantic graph 1 for each other request so we start at the

11:55

very finest level of detail and we want to merge the nodes that are close in the sense of travel time and you so 1 way to view this problem is to say that we have this here we have a given cluster size they want to work on the level of a 10 minute walking distance for example and I and how can I find the the cluster distribution that requires the least number of clusters and it turns out that this question is I computationally more complex it's NP-hard so but there are approximations that thorough efficiency so so that's what we do of the soul essentially this this way of doing it requires 1 1 single source all shortest path complication per cluster where and the number of nodes that enter the each computation is determined by the variational how many how many clusters you want per units area for example so this is just an example again of clustering the notes in the previous example are shown in different colors only the 10 minute walking distance so cost functions of 1st we derive cost functions for what we call a standard conditions differentiating between roads power and terrain and the standard cost the functions are used for the generating the grade which are which is at the base of the application and then I'll for the shortest path request we we compute weights dynamically so the standard cost functions are based on literature sociology literature and more a little bit of physics and during cold data and then scale factors can be applied to account for the effect of uh same moist ground state round of snow roughness roughness tends to tends to decrease the speed with a few % etc. so that this is kind of of has to be tuned or learned ideally based on actual empirical data and we have different categories for a vehicles bicycles hikers and so this is an example of a standard cost function for a for a hybrid and maximum speed here is about at about the 3 degree downward slope I think and if you have a cost function for energy the most efficient I believe through receptive of the 10 degree downward slope so this is speed what we need is actually this long of the inverse of speed to In this equivalence and so the system is so consists of a 0 bps server other the graph historian cost grows with the post GIS extensions so we use the reading algorithms in this case the area start to waste our time and so it as an example of a nested BPS request we can use uh a shortest path at the coarse graph level as input to the shortest path a request at the next level etc. This is naturally implemented as a nested bps request of course you can do it otherwise so straightforward so this is client that my colleague work announced and implemented we currently have something like 30 to 40 of these once the services for various applications on ship traffic monitoring is 1 example from satellite image analysis and Bayesian analysis so this is the 1st

16:36

routine example what you smaller red flags marking the start and the end of the and and this is a this is an observer placed in the center of the image it's field of view and we want to avoid it so that affects the outcome of the reading of the all of the optimal route if you take away the observer this is the shortest path this is a topographic map it's about so it shows that the area is quite hilly on the next example is the example of hierarchical iteration in a very coarse grid the 1st solution to the 2nd and the final 5

17:37

so what are the next steps for us is to elaborate on the ground physics simulations which is really the region that office does the simulations such malls are used as as boundary conditions to weather forecast models but they can also be run independently offline as they say would perhaps more detailed background physics models and they're quite useful for predicting so that temperature profiles and which also implies that something about the load carrying capacity of the ground

18:21

of so the summary is that we we try to do path planning to rain it's still a work in progress I should say we have a situation dependent edge weights we use a hierarchical random graph service-oriented implementation WES and and post GIS system important complement for storing the graph with PG rooting uh shortest path algorithms on a so we we're on our

18:55

work heavily on the open source software so it's just to acknowledge that and the the final slide in

19:07

my my colleagues putting a joke here I don't know if anyone can spot but it's still a 0 lord will the we have a lot of use in Norway the the but been duped that few indeed questioned 1 of the and Hamilton environments of an enlightened but actually that's what we chose this area good lot to you wanted to test the algorithm in an area we know well so the and we also doing this portable orienteering so well known in the US but so we're all of whom had of that that's why you chose this this is and Northern side also all the time for the creation it any other questions think so so thank you very much

00:00

Resultante

Web Services

Prozess <Physik>

Automatische Handlungsplanung

Kartesische Koordinaten

Partikelsystem

Kombinatorische Gruppentheorie

Synchronisierung

Computeranimation

Standardabweichung

00:54

Knotenmenge

Punkt

Gewicht <Mathematik>

Graph

Datennetz

Punkt

Wort <Informatik>

Routing

Automatische Handlungsplanung

Computeranimation

01:31

Schätzwert

Task

Wellenpaket

Flächeninhalt

Graph

Schätzung

Automatische Handlungsplanung

Punkt

Kartesische Koordinaten

Routing

Überlagerung <Mathematik>

Automatische Handlungsplanung

Computeranimation

02:20

Bit

Gewicht <Mathematik>

Punkt

Physikalismus

Automatische Handlungsplanung

Implementierung

Kartesische Koordinaten

Information

Textur-Mapping

Computeranimation

Überlagerung <Mathematik>

Spezialrechner

Graph

Prognoseverfahren

Perspektive

Theoretische Physik

Diskrete Simulation

Datentyp

Implementierung

Serviceorientierte Architektur

Bildauflösung

Perspektive

Graph

Datenmodell

Profil <Aerodynamik>

Kanalkapazität

Automatische Handlungsplanung

Quick-Sort

Arithmetisches Mittel

Generator <Informatik>

Last

Information

Simulation

Beobachtungsstudie

03:53

Distributionstheorie

Ortsoperator

Graph

Kategorie <Mathematik>

Zufallsgraph

Voronoi-Diagramm

Polygonnetz

Regulärer Graph

Zellularer Automat

Zahlenbereich

Polygon

Roboter

Graph

Knotenmenge

Zufallszahlen

Regulärer Graph

Spieltheorie

Typentheorie

Datentyp

Wort <Informatik>

Diagonale <Geometrie>

Gerade

Einflussgröße

06:07

Wechselsprung

Datennetz

Typentheorie

Datentyp

Varianz

Normalvektor

Menge

Varianz

Computeranimation

Überlagerung <Mathematik>

Attributierte Grammatik

Gradient

06:41

Wald <Graphentheorie>

Sichtenkonzept

Momentenproblem

Datennetz

Kategorie <Mathematik>

Physikalismus

Polygon

Computeranimation

Weg <Topologie>

Polygon

Menge

Flächeninhalt

Digitale Photographie

Datennetz

Typentheorie

Datentyp

07:50

Zentrische Streckung

Graph

Dichte <Physik>

Selbstrepräsentation

Zufallsgraph

Relativitätstheorie

Zahlenbereich

Unrundheit

Knotenmenge

Auflösungsvermögen

Computeranimation

Übergang

Dichte <Physik>

Graph

Knotenmenge

Zufallsgraph

Mathematische Morphologie

Selbstrepräsentation

Attributierte Grammatik

Vorlesung/Konferenz

Triangulierung

Attributierte Grammatik

09:06

Konstante

Knotenmenge

Gewicht <Mathematik>

Gewicht <Mathematik>

Graph

Knotenmenge

Bildauflösung

09:49

Matrizenrechnung

Gewicht <Mathematik>

Graph

Matrizenrechnung

Element <Mathematik>

Schwach besetzte Matrix

Knotenmenge

Computeranimation

Schwach besetzte Matrix

Graph

Knotenmenge

Gewicht <Mathematik>

Flächentheorie

Selbstrepräsentation

Diagonale <Geometrie>

10:31

Satellitensystem

Distributionstheorie

Bit

Polygonzug

Selbstrepräsentation

Kartesische Koordinaten

Computerunterstütztes Verfahren

Rechenbuch

Analysis

Computeranimation

Übergang

Gradient

Gotcha <Informatik>

Kürzester-Weg-Problem

Client

Web Services

Algorithmus

Einheit <Mathematik>

Maßstab

Standardabweichung

Konditionszahl

Bilderkennung

Einflussgröße

Inklusion <Mathematik>

Zentrische Streckung

Lineares Funktional

Approximation

Kategorie <Mathematik>

Inverse

Quellcode

Ein-Ausgabe

Knotenmenge

Teilbarkeit

Hierarchische Struktur

Funktion <Mathematik>

Konditionszahl

Client

Kategorie <Mathematik>

Standardabweichung

Lesen <Datenverarbeitung>

Partitionsfunktion

Gewicht <Mathematik>

Physikalismus

Zahlenbereich

Feuchtigkeit

Unrundheit

Maßerweiterung

Physikalisches System

Open Source

Graph

Knotenmenge

Gewicht <Mathematik>

Theoretische Physik

Kostenfunktion

Schätzung

Kommunalität

Abstand

Cluster <Rechnernetz>

Maßerweiterung

Analysis

Leistung <Physik>

Soundverarbeitung

Algorithmus

Graph

Bildauflösung

Physikalisches System

Automatische Handlungsplanung

Inverser Limes

Abstand

Energiedichte

Bildschirmmaske

Minimalgrad

Flächeninhalt

Loop

Selbstrepräsentation

Kantenfärbung

Klumpenstichprobe

Greedy-Algorithmus

16:34

Physikalismus

Iteration

Computeranimation

Informationsmodellierung

Fahne <Mathematik>

Diskrete Simulation

Theoretische Physik

Luenberger-Beobachter

Vorlesung/Konferenz

Bildgebendes Verfahren

Normalvektor

Hierarchie <Mathematik>

Sichtenkonzept

Kanalkapazität

Profil <Aerodynamik>

Routing

Übergang

Office-Paket

Mapping <Computergraphik>

Benutzerprofil

Randwert

Datenfeld

Flächeninhalt

Last

Einheit <Mathematik>

Lesen <Datenverarbeitung>

18:20

Web Services

Hierarchie <Mathematik>

Offene Menge

Gewicht <Mathematik>

Graph

Open Source

Zufallsgraph

Implementierung

Physikalisches System

Automatische Handlungsplanung

Computeranimation

Graph

Hierarchische Struktur

Algorithmus

Zufallsgraph

Gewicht <Mathematik>

Arithmetische Folge

Software

Selbstrepräsentation

Implementierung

Serviceorientierte Architektur

19:05

Orientierung <Mathematik>

Algorithmus

Flächeninhalt

Güte der Anpassung

Vorlesung/Konferenz

Programmierumgebung

Computeranimation

### Metadaten

#### Formale Metadaten

Titel | The way to go with WPS |

Serientitel | FOSS4G Seoul 2015 |

Autor | Messel, Espen |

Lizenz |
CC-Namensnennung - keine kommerzielle Nutzung - Weitergabe unter gleichen Bedingungen 3.0 Deutschland: Sie dürfen das Werk bzw. den Inhalt zu jedem legalen und nicht-kommerziellen Zweck nutzen, verändern und in unveränderter oder veränderter Form vervielfältigen, verbreiten und öffentlich zugänglich machen, sofern Sie den Namen des Autors/Rechteinhabers in der von ihm festgelegten Weise nennen und das Werk bzw. diesen Inhalt auch in veränderter Form nur unter den Bedingungen dieser Lizenz weitergeben. |

DOI | 10.5446/32093 |

Herausgeber | FOSS4G |

Erscheinungsjahr | 2015 |

Sprache | Englisch |

Produzent | FOSS4G KOREA |

Produktionsjahr | 2015 |

Produktionsort | Seoul, South Korea |

#### Inhaltliche Metadaten

Fachgebiet | Informatik |

Abstract | How to find your way in difficult terrain, with obstacles, hazards, and deep snow? We present a solution for cross-country path planning and mobility, based on OSGeo software and open data. A large graph representing terrain, roads, and paths is stored in PostGIS for use with the pgRouting module of shortest path algorithms. The graph is based on detailed topography, soil type and vegetation data, and edge weights can be adapted for hikers and vehicles. The application is service oriented and held together by the Web Processing Service (WPS), the OGC interface standard for computation-oriented web services. A key component is the ZOO WPS server. The presentation will discuss WPS benefits and describe graph and weight generation, including challenges such as accounting for dynamic data about temporary hazards, weather, etc. |