Merken

Mapping the world beyond 3857

Zitierlink des Filmsegments
Embed Code

Automatisierte Medienanalyse

Beta
Erkannte Entitäten
Sprachtranskript
but
and scalable
so 5 sets uh so they're free you have no financial commitment to you of several version of them to choose from so of course it's a it's a very easy starting point for your mapping brother who have a background map of which is rich why because it's
funny everybody knows Google Maps everybody knows OpenStreetMap and so on and uh we tend to see the word that this way which is a bit of a tragedy my opinion cos Antarctic and Greenland are not that big but anyways that's how people are used to to see the work today
why because it's it's easy uh there are tens of interesting map projections out there representing the word in different ways but why you do you have to learn all of them are there if you can just get away with 1 so it's convenient it's easy you can relax In the has another bit since
we are closer to they they claim that I used to being I think that everybody in New Zealand would be pretty happy about this projection nicely wrapping around them of showing the Pacific Ocean is a solid there uh seem less uh entity instead of you know cutting its uh at the edges so it's also nice because it's very to the dateline
so if why am I talking today why am I E had them you're talking about projections outside of 38 57 where there is a a few reasons why people do not want to use good Mercator 1 reason to rule them all
its distortion we use map projections for 150 cannot reason to minimize the distortion ascertain type of distortion in ascertain area the recent changes depending on what type of distortion are we talking about shape distortion area distortion distances between points in that we changed between the areas that we're trying to look at how we trying to map the global state local scale or maybe we are bound by legal requirements like for example in Italy you have to publish in you can petition or than 3 North that's the law then you can use all the projections if you want but those 2 Africa so if you're publishing legally binding maps you have to use those not the if the 57 you can also add to the 71 so let's have a look at the
the area distortion problem the anything that's close to the pool gets inflated lot if he thought it was moved the more or less where getting land is you would be much much bigger but then again it's way colder there so I'm I'm not proposing to move
um how what we do in case we want our my good representation of the area of their land masses and the scientists and their relative sizes we use equal area projections of this is very important for statistical maps and for didactical purposes so that people don't think that the Antarctic ice the biggest Compaq continent in the war I do here is 1 example
of some over and the aquarium at this is the more weight projection and and feel of the Food and Agriculture within each each iteration sponsored you solutions to implemented in just over 4 of their atlas of Fisheries in which they have a big polygons over the earth and we need to have them show that I proportions so you you can actually see what's began what a smaller without having to think all it's not so if the but I have to think it's a smaller so and
2nd distortion distance perception of yes of most of the tools that we have out there you have a point and click to some measure tool that you can click a path and get a distance but the problem is that you cannot shut down your eyes and your eyes with people seeing a relative distances reminder level to tell you it's 1 thousand kilometres but he might be able to tell you 0 it looks like it's twice as long as the all other direction and that you know what Mercator had their the distances are all wrong this is a map of over lines of equal distance from a point in Slovenia and as you can see they are not circles they look like cats instead so your perception of distances on lubricator is completely wrong so what do we do if we
need actual préciser representation of relative distances we use of the equity spent months instead antiquities does not use a map in which certainer lines are proportional to the directions it's not possible to make our projection that's on operate in terms of the censors everywhere but for example these of the novel I Could be sh equidistant map shows the right uh so that i the distance from the pole to any other point in the war so uh acquittances maps are important when you have a point and you want to show the distances to all other points around the and then you might be familiar with it because the United Nations Global is an elegant derived from an equidistant at the metallic instance productions here is 1 use case from our
canadian weather map um they have our weather station and they are displaying everything that around it in as you call equidistant projection so they basically have 1 different map for each of weather station so that they can uh literally show through the senses uh from that article point which is having finally uh they would
good Mercator is not showing up bowls not not at all you can see a bit on article but you're not seeing it all in the north pole is also cut out and yet also very important for the future of humanity for climate studies and so on so there must be a way to map them correctly and show them in a manner and the yes that way is the
posturographic commonly at least it's the posturographic projections from this is the 1 example from the British Antarctic Survey which is powered by just
uh the this story this is another
1 it's called Paula view it's showing the number of climate-related datasets and the poles this 1 shows both and particle in the north pole is up with a view of the north pole and the SCO itself finger
that fisheries outlast allows you to look at the war the from different projections to appreciate the different aspects of it in 1 of them is there a self for so called on the graphic again powered by Justice
yeah and 38 57 were American is actually the besides the north and the south area's actually quite well behaved what do we try to display global datasets with all these projections well we take each point that we have in the
vector same each pixel that we have in the rest of datasets we project them point by point and should be all final no well no not exactly know they can be very rough it can go very wrong and I want to show you some examples of what happens if you have a
global dataset and you try to display with local projections without preprocessing it without cutting it without selecting it just as it gets which is actually a more common use case that you can might think of so this is due to just meet binding and
trying to look at the Pacific and I would really like to see America's there'd but they are not there because I'm asking for an area which is between a hot uh 0 degrees longitude and the something like 400 degrees longitude and there's nothing be beyond 180 solely from just asking the database give me something that is beyond the dateline is gonna give me back nothing so I have a big hole in the map which is not exactly knots what if I try it with a projection which is designed to to look at the Pacific Ocean such as the Pacific American city barometric 1 well Antarctica goes away because it touches the pole and American rejections brawl blows up on the pool of degenerates to infinity and the greenlandic was wrapped uh in such a way that it seems it's covering the entire north pole and that's also problem shouldn't be that way what if I try to use a Lambert
Conformal conic will some of their countries are being stretched in a funny way not put uh in the upper part of the MAP what's that shouldn't be there I I will show you later how we should look like
what if I try to be to do something so in a sense such as a to change I I'm not even change in projection just changing the ellipsoid from over geographic coordinates geographic coordinates are they can shift the changes of the of the value of the geographic coordinates and those that are close to the dateline microsecond and on the other side so I end up with the long line crossing my maximum wondering what's going on in the
world this this this comment itself is is you the 32 north over Italy trying to reproduce the whole war the best that is altogether if you try to just a project point like and not exactly what you want so what's
going on here well we have a global dataset and we are trying to use local defined locally defined projections so we should really filter the daytime probably cut it causes some of the polygons of so huge that they go far away from the projection and the area of information then we have data that is crossing the dateline after very projection because of the shoes and stuff like that and so we have those long lines and we should be cutting the date of the lives of 8 lines instead uh and then we have the problem of the long lines like me a median or Apollo we I only have 2 points if I reported 2 points but that the output is supposed to be a curve I have a problem because if I refer 2 point I end up with 2 points I should intensify before product yeah where we have to do all of these by hand them I'm tired just about describing so uh just server our weight and it gets worse in Dublin as there are the older projections the all the projections are a way to get your own coordinate the reference system service center in your point of choice instead of using 1 of the 5 thousand already well known projection in is due to the things so let's say you want uh them all way that centered on courtier I can use in your projections the thing is anybody can give me a different production center so basically have an infinite number of projections that I could theoretically handle how is this this is a mass uh and uh it's impossible to compute by hand in the use cases in which you have many many projections in the output that you want to handle it needs to be automated induce ServerJet does just that with a system that we call the advanced projection and there they are a smart classes that know the issues and the problems of each of 1 of the projections and can uh fix them before why we try to paint we have the but as most of them from the most common projections and it's growing over time so you can see I have a list we have 1 refers directly performed by metadata of the Lambert for the colony care for posturographic and so on so how do
does this work well it's 4 steps at the moment the first one is you gave me an area that you want to display i have to figure out where the source data for it years and sometimes it's not what you lost you ask for a bounding box that's crossing the declining going outside of the body of a range of longitudes well in that case I also have to read from the other end of the war so I'm trying to display the yellow area but they have to be wary of the 2 adopted areas that so the system knows about this and we read all the data the uh then I
have to cut excess data if I'm trying to display our UTM projections there's no way I can reproject data which is too far away from the central meridian strive to cut away it's of selection in particles according tipping the coast I denote the it's too big I have likely cut it it's part of my men but most part of it is clear and then I have
to write the data I read it from 1 side of the world and they have to move it In the area that was requested so I'm taking the data and translating so that it shows up in my map as a seamless problems finally for all
those cases in which the re-projection calls the points to wrap around like then generate very long line crossing the work that I have to leave out all their results so that it shows up and I part of the world and and duplicate people on the other side in carries on looking at the whole time but so let's try
again the same as before but with a vast projection and link enabled which is by the way enabled by default service so you don't have to do anything you might want to disable it uh but non enabled by the
so know this is the map on Pacific and this time I have the americas of showing why the Coase the system the 1st that I was reading beyond the dateline went and query the other side of the war that took it and pushed it in the right place so it's generated out of this is a single of mess request OK so they're not fighting x here so the system actually and took uh took 3 thousand movie can place this is the Mercator that I was showing before so you can see now everything is uh proper nicely land is complete and the Antarctica is there because I got it the projection and the format of medieval uh knows that we don't want to to show anything beyond 85 degrees north and a or below 85 degrees south so it just cuts Texas This is the Lambert
Conformal conic no more funkiness of the bar at the top this is how what a conformal conic should show there's nothing at all because we we brought the year with a cold and then we open so if you think about opening the call that's what you get so this is that they can change again
know of funky along lines of crossing the because this system effectively there uh big crossing and fixed at the rear of the data across the lately
so this is what you we have today it's working for vector data has shown that it's also working for us that they can starting we just over 2 7 so 6 months ago what's cooking what we're working on right now while
many scientific data sets have a legal problem they don't have coordinates between minus 180 in 118 onto the for some reason they choose to both between 0 and 360 and when you put that into our system that you know once right the longitudes the doesn't like it and so much so we are utterly fixing the rendering subsystem in order to recognize that situation and handle it properly so that all the properties of their vast projection and links are preserved when the dataset uh is between 0 and 360 come and then there
was the problem of the long lines that I was citing before I have no 1 you or 1 parallel represented of a single line between 2 points I to the both the points and they get back again a straight line but some projection current long lines and curves so I need to densify you need to apply the points before sort of the curves show that the problem is how much I don't know depends on what kind of projection what kind of area so it's something that I would need to decide dynamic we have a bit of cold in there was there a projection subsystem that does this kind of analysis to determine whether we can re project that ask the as an affine transform or a piecewise affine transform their of in my spare time I'm I'm and trying to leverage it to figure out how many extra points and plot and the long lines in order for the curves to seems to show up in the universe of but it's not you finally uh as a
closing water uh and we have seen no they're just during the last 6 months uh I don't know 4 or 5 new production showing up in the duties called ways contributed by people so I'm pretty sure that there is a lot of interest for us to uh mark projections outside of Web Mercator either for reading the power for producing of that's are we have seen in the generation of new auto quotes for polar stereographic and mnemonic so people definitely want to the center there only the projections where they want without applying to the fused you database so I'm pretty sure they would be more need for a large collection and link in the future I so I can I can state for sure that there is life outside of the key fits in with
this I am but questions yes who it was a and but on the Julian when they come back together too weak to be this that all those back together now that's a possibility but that the problem is that he would introduce those Oslo topological operation in and saying and I'm not to have the uh to to go there if there there is funding and if someone wants an option to enable them we could have
next yeah the Antarctic ice as the
little black line in the middle we could change it so that that little black line doesn't appear anymore because the Chu Mbits get union before showing them but is a set up at the moment it would be pretty expensive so I I really don't want to go there uh well actually if you look at Russia there's also a line containing and that that's the same problem I the right yeah we would also after our way to identify the different bits and merge them together so I mean it's complicated it would be nice no question but did you wanna start and so I understood
you know this from the user's group to say that most of it we just over 2 7 uh arranges rich wake you get the summer next that I improvements but so the vector Bayes that uh advanced projection Memling is actually for 5 years old since the 2012 not 2010 the 1st version and then we treat the improved it added new projection support and so on and just over 2 7 we added the support for us to which was not enjoying a bounce projection and link support so if you were taking a war the maps and showing it in the forest Eurocratic there was a larger why that was the short arm and then into 8 we increase the number of projections we support and made some tweaks uh for the 0 to 360 part of it is writing their into weight so it is moving on it improving over time all right and the interesting and useful and and so on and
Mapping <Computergraphik>
Punkt
Menge
Java Enterprise
Client
Versionsverwaltung
Familie <Mathematik>
Applet
Bildverarbeitung
Visualisierung
Textur-Mapping
Datenfusion
Hinterlegungsverfahren <Kryptologie>
Computeranimation
Mapping <Computergraphik>
Subtraktion
Bit
Zehn
Affine Varietät
sinc-Funktion
Wort <Informatik>
Affine Varietät
Benutzerführung
Computeranimation
Zentrische Streckung
Stellenring
Shape <Informatik>
Punkt
Affine Varietät
Mathematisierung
Stellenring
Gesetz <Physik>
Verzerrungstensor
Computeranimation
Mapping <Computergraphik>
Abstand
Textur-Mapping
Flächeninhalt
Datentyp
Flächeninhalt
Verzerrungstensor
Abstand
Shape <Informatik>
Aggregatzustand
Textur-Mapping
Statistik
Gewicht <Mathematik>
Flächeninhalt
Affine Varietät
Differenzkern
Selbstrepräsentation
Azimut
Iteration
Ruhmasse
Flächeninhalt
Gleichheitszeichen
Polygon
Computeranimation
Punkt
Kreisfläche
Affine Varietät
Selbstrepräsentation
Relativitätstheorie
Biprodukt
Term
Übergang
Richtung
Mapping <Computergraphik>
Textur-Mapping
Polstelle
Differenzkern
Verzerrungstensor
Abstand
Computerunterstützte Übersetzung
Gerade
Instantiierung
Beobachtungsstudie
Mapping <Computergraphik>
Polstelle
Bit
Punkt
Affine Varietät
Arbeitsplatzcomputer
Polstelle
Beobachtungsstudie
Distributionstheorie
Affine Varietät
Polare
ART-Netz
FAQ
Sondierung
Computeranimation
Distributionstheorie
Polstelle
Sichtenkonzept
Affine Varietät
Polare
ART-Netz
Zahlenbereich
Partikelsystem
FAQ
Sichtenkonzept
Computeranimation
Open Source
Pixel
Punkt
Affine Varietät
Flächeninhalt
Axonometrie
Vektorraum
Computeranimation
Maschinenschreiben
Knoten <Mathematik>
Affine Varietät
Datenhaltung
Entartung <Mathematik>
Computeranimation
Unendlichkeit
Mapping <Computergraphik>
Polstelle
Minimalgrad
Flächeninhalt
Räumliche Anordnung
Ganze Funktion
Mapping <Computergraphik>
Kegelschnitt
Affine Varietät
Extrempunkt
Kegelschnitt
Mathematisierung
Mereologie
Gerade
Subtraktion
Decodierung
Gewicht <Mathematik>
Punkt
Klasse <Mathematik>
Zahlenbereich
Extrempunkt
Polygon
Verzerrungstensor
Computeranimation
Metadaten
Kurvenanpassung
Gerade
Auswahlaxiom
Funktion <Mathematik>
Kugel
Affine Varietät
Kegelschnitt
Ruhmasse
Mailing-Liste
Physikalisches System
Objektklasse
Biprodukt
Medianwert
Menge
Unendlichkeit
Dienst <Informatik>
Beamer
Flächeninhalt
Information
Prozessautomation
Spannweite <Stochastik>
Affine Varietät
Unified Threat Management
Quader
Momentenproblem
Flächeninhalt
Trennschärfe <Statistik>
Mereologie
Lesen <Datenverarbeitung>
Partikelsystem
Physikalisches System
Quellcode
Schnitt <Graphentheorie>
Computeranimation
Bitmap-Graphik
Resultante
Mapping <Computergraphik>
Punkt
Flächeninhalt
Mereologie
Lesen <Datenverarbeitung>
Gerade
Computeranimation
Mapping <Computergraphik>
Dienst <Informatik>
Minimalgrad
Affine Varietät
Axonometrie
Physikalisches System
Binder <Informatik>
Default
Computeranimation
Kegelschnitt
Kegelschnitt
Systemaufruf
Physikalisches System
Gerade
Computeranimation
Affine Varietät
Menge
Kategorie <Mathematik>
Rechter Winkel
Physikalisches System
Vektorraum
Ordnung <Mathematik>
Binder <Informatik>
Computeranimation
Decodierung
Punkt
Wasserdampftafel
Transformation <Mathematik>
Computeranimation
Quellcode
Benutzerbeteiligung
Affiner Raum
Punkt
Kurvenanpassung
Gerade
Analysis
Leistung <Physik>
Affine Varietät
Datenhaltung
Axonometrie
Strömungsrichtung
Plot <Graphische Darstellung>
Gasströmung
Binder <Informatik>
Biprodukt
Quick-Sort
Generator <Informatik>
Flächeninhalt
Polarisation
Ordnung <Mathematik>
Netzwerktopologie
Nichtlinearer Operator
Kegelschnitt
Computeranimation
Konfiguration <Informatik>
Bit
Wald <Graphentheorie>
Gewicht <Mathematik>
Momentenproblem
Affine Varietät
Gruppenkeim
Versionsverwaltung
Zahlenbereich
Vektorraum
Binder <Informatik>
Computeranimation
Open Source
Textur-Mapping
Mereologie
Gerade
Vorlesung/Konferenz

Metadaten

Formale Metadaten

Titel Mapping the world beyond 3857
Serientitel FOSS4G Seoul 2015
Autor Aime, Andrea
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 10.5446/32036
Herausgeber FOSS4G
Erscheinungsjahr 2015
Sprache Englisch
Produzent FOSS4G KOREA
Produktionsjahr 2015
Produktionsort Seoul, South Korea

Inhaltliche Metadaten

Fachgebiet Informatik
Abstract Most popular mapping presentations today, ranging from clients to servers, show and discuss only maps in EPSG:3857, the popular Mercator derived projection used by OSM as well as most commercial tiles providers. There is however an interesting, exciting world of map projections out there, that are still being used in a variety of context. This presentation will introduce the advancement made in GeoTools and GeoServer to handle those use cases, where users have a worldwide data set, and need to view all or part of it in multiple projections, some of which valid in a limited area, and requiring the software to perform a proper display of it on the fly, without any preparation. We'll discuss GeoTools/GeoServer advanced projection handling manages to deal with these cases, wrapping data, dealing with the poles and the dateline, cutting on the fly excess data, densifying on the fly long lines as needed to ensure a smooth reprojection, for a variety of cases, ranging from seemingly innocuous datum shifts, maps having the prime meridian over the pacific, and the various tricks to properly handle stereographic, transverse mercator, Lambert conic and other limited area projections against world wide source data sets.

Ähnliche Filme

Loading...