Bestand wählen

Automatic image registration on 3D-to-2D images

Zitierlink des Filmsegments
Embed Code

Automatisierte Medienanalyse

Erkannte Entitäten
and we have no that 1 or that language is running it will be presented by the major from a high performance computing
research and and development division of the National Electronics and Computer Technology Center in Bangkok the title is automatic image registration between object and in 18 eighties good morning everybody I have been in online and thinking about the and have to take from the poem that and it became gold mine topic from the of a mistake in actually taken from 3 D to 2 2 the be better now to other people who work mention of the fact the basic idea of the paper is that basically imagine you create any of the process of transforming of multiple dimensions to within an image on the family that the support the in general if the thought opportunity come from the same like the nation up at like at applying that under the site and the right side he he that is the data they did you that he people locations and applying transformation and finally we can get the the reduce the internet and the final product if the data come from different parts of the predictions like at that site is from and 1 is scanning the right side from them and therefore the on the left side of the plural prediction right right in the center of attention to from different prediction how to do on this 1 quality but that's the way to do all of my paper they actually are the weighted has gone from then protection to Apollo predicted it be that to alter rectification but then the problem of ontology that could be taken it the complex and methodology the 1st and 2nd it had to rely on the and we have to do something about it in that configuration that work for all at the report of that
that really so I ordered paper attracted that that attracted a lot to improve model 2 3 different costs complications if if the proposed methodology this that from the mental type a lot of histamine it's all you remember time points and then you go through that type of selection bit confirmation model here I used the polynomial model and then you go through the automated increasing type like that's on and the improvement of the accuracy values in the corner and finally we do accounting on that data but we have to to demand a tripod what our that's the on uh the edge you can get some some point the devise some complied by many because the data input the right in the corner that he not the real we'll calling and I use the coordinates from the for the shop that picture by the picture by type reflection we implied a confirmation moral confirmation model yet we modified something and they can be in there them a model thought to be climate of x y and it had more out of 20 if you want you a vacuum yes but that it's so complex for me I try to reduce that 1 it is the chart formal polynomial I used for and this paper and and we try to compute by group what that thought and also rely on the adhesion feeding them to kick and equity it is that we when we try to to to measure the combination of the output we found that the time points of input data you got correlation in go why what it a control around 0 . 9 9 and 0 . 9 5 in the x and column and when the type might include to can type the correlation is better than the previous 1 and then there are it is not their output the have from the object that in and in the middle image that that 1 is out on the output of a transformation that they are not beautiful but you do something on that 1 is not OK there are let's talk about economy increasing type the previous slide we can see that the the the the beautiful not good so the problem that come from the time points so small that 9 or 10 how to look at now the thing that how to increase the number of type by politically not governmental it it's related to automatically on on the on increasing the type you can't do that on the edge of the know and transformation model we can find out the the type of life and the matched points between the mass that he meant and fled the mansion and although I tried to do well on the corner they're on the template defined templates you can get from corner this means that now I have some point at some time point from the BH and all of those some type of pipeline from the corner I have to refer bet on the whole application I can get funding from that 1 the foot of the G. D. P. K. A. you and today it cats the Ugandan point I mean all applied based on data that can be known In 1 when I created that the window 2 because I can go around the pin and then finally the King Fahd created the idea of a mean of 13 and the if you take a look under the correlation of time points we found that when the number of time points this increase the correlation of the output is also increase it is the result of the revamping and I think that we get from this paper is that but it makes you do all that the output is around to pixels this is the accuracy that
I got from from this that the when we applied for all of the beach area other people it is the product of the transformation of the paper conclusion here I consider
that the bet on the ability you can you can get a great PR tool and that that he'll and also the this 1 can be applied all they there that experiments because you need to do is take a mapping on that data this is the way that that I think that very good for class actually even this 1 is not even then on graphs like that kind of something that think that in the the further work we can do extent this 1 to crash thank you that might OK
comments or questions but I had but if you have 2
images of this kind of thing and you want to know the register then what happens if 1 images taken during the day bright light and the other 1 is taken in and shake but you applied I tell you that make great energy because it might lead you on the spot and up if you matching all registering 2 images of the signs saying but in your
case in 1 image is taking the bright
light about like that and the other 1 is taken
up in a Shiite situation how can you have registered that you apply some preprocessing the light no no I have no idea that 1
I haven't I fall fall for the intensity value that people will be different but at end each the exactly designed but intensity that will be different so therefore the matching will have some problem that I not there OK sorry no idea of different that the right that only want to inform you that the that that
photograph of during the programs to make economic photos that means the can all go from all direction of 1 plane and they put them together approach to make it three-dimensional photo and I think you kind of the lecture so your approach you can the through the program the front and the pitch is 1 of the program but I and looking from whole they admit sharing educational and indicated
you go on the mentioned to me that meant that we might obtained OK and then there's that if this moral sense
him there there and there will reconvene yeah
at half past 2 OK you this could be for everyone who has presentation on the uh to the gamma and you would be a value OK and then the posterior
conditioned to this the I
Formale Sprache
Prozess <Physik>
Spiegelung <Mathematik>
Familie <Mathematik>
Formale Grammatik
Kartesische Koordinaten
Komplex <Algebra>
Trennschärfe <Statistik>
Funktion <Mathematik>
Güte der Anpassung
Arithmetisches Mittel
Rechter Winkel
Registrierung <Bildverarbeitung>
Computerunterstützte Übersetzung
Web Site
Transformation <Mathematik>
Data Mining
Bildgebendes Verfahren
Ontologie <Wissensverarbeitung>
Persönliche Identifikationsnummer
Objekt <Kategorie>
Mapping <Computergraphik>
Klasse <Mathematik>
Ungerichteter Graph
Transformation <Mathematik>
Vorzeichen <Mathematik>
Bildgebendes Verfahren
Rechter Winkel
Digitale Photographie
Kombinatorische Gruppentheorie


Formale Metadaten

Titel Automatic image registration on 3D-to-2D images
Serientitel Open source GIS - GRASS user conference 2002
Anzahl der Teile 45
Autor Tipdecho, Taravudhe
Lizenz CC-Namensnennung - keine Bearbeitung 3.0 Deutschland:
Sie dürfen das Werk in unveränderter Form zu jedem legalen Zweck nutzen, vervielfältigen, verbreiten und öffentlich zugänglich machen, sofern Sie den Namen des Autors/Rechteinhabers in der von ihm festgelegten Weise nennen.
DOI 10.5446/21768
Herausgeber University of Trento
Erscheinungsjahr 2002
Sprache Englisch

Inhaltliche Metadaten

Fachgebiet Informatik

Ähnliche Filme