ILMSImage Classification

From ILMS-Wiki
(Difference between revisions)
Jump to: navigation, search
(Unsupervised Classification in Cell Type Classes)
m
 
(27 intermediate revisions by 2 users not shown)
Line 1: Line 1:
=[[Bild: ilms_img_classification_icon.png|50px|<span title=""></span>]] ILMSImage Classification=
+
[[de:ILMSImage_Classification]]
 +
[[pt:ILMSImage_Classification]]
 +
=[[File: ilms_img_classification_icon.png|50px|<span title=""></span>]] ILMSImage Classification=
  
 
==Introduction==
 
==Introduction==
  
''ILMSImage Classification'' is part of the [[Tutorial_ILMSImage#ILMSImage_Plug-in_for_QuantumGIS|ILMSImage plug-in for QuantumGIS]] and in this context it carries out a two-step thematic analysis, comprising unsupervised and supervised classification. u
+
''ILMSImage Classification'' is part of the [[Tutorial_ILMSImage#ILMSImage_Plug-in_for_QuantumGIS|ILMSImage plug-in for QuantumGIS]] and in this context it carries out a two-step thematic analysis, comprising unsupervised and supervised classification.  
  
Like various ILMSImage panels it consists of two components, the ''[[ILMSImage_Cell_Creation_english#Part_1:_Information|ILMSImage Project Information]]'' in the upper section and the actual tools in the lower section.
+
Like various ILMSImage panels it consists of two components, the ''[[ILMSImage_Cell_Creation#Part_1:_Information|ILMSImage Project Information]]'' in the upper section and the actual tools in the lower section.
  
[[Bild:ilms_img_qgis_work_16.png|Das ILMSImage-Panel zur thematischen Klassifikation]]
+
[[File:ilms_img_qgis_work_16.png|Das ILMSImage-Panel zur thematischen Klassifikation]]
  
 
==Thematic Classification==
 
==Thematic Classification==
Line 16: Line 18:
 
and (3) the actual thematic classification which is based on results of both preceding tasks. Below every step is presented and described in detail.  
 
and (3) the actual thematic classification which is based on results of both preceding tasks. Below every step is presented and described in detail.  
  
===Unsupervised Classification in Cell Type Classes===
+
===Unsupervised Classification into Cell Type Classes===
  
The unsupervised classification refers to generated geometries during the [[ILMSImage Cell Creation english|cell creation]] and the  [[ILMSImage Attribute Calculation english|calculated attributes]] for the geometries to derive cell type classes from both components. They are group of cells which - no matter which geographical location in the image or location to one another - have similar features. Simply put, in this task e.g. elongated and dark cells are separated from those which are rather round and bright - only that the corresponding decision in reality is not limited to the named features but is based on all features generated during attribute calculation. The basic concept of this task therefore corresponds to the partitional cluster analysis ([http://en.wikipedia.org/wiki/Cluster_analysis#Partitional_clustering| Partitional clustering]).
+
The unsupervised classification refers to generated geometries during the [[ILMSImage Cell Creation|cell creation]] and the  [[ILMSImage Attribute Calculation|calculated attributes]] for the geometries to derive cell type classes from both components. They are group of cells which - no matter which geographical location in the image or location to one another - have similar features. Simply put, in this task e.g. elongated and dark cells are separated from those which are rather round and bright - only that the corresponding decision in reality is not limited to the named features but is based on all features generated during attribute calculation. The basic concept of this task therefore corresponds to the partitional cluster analysis ([http://en.wikipedia.org/wiki/Cluster_analysis#Partitional_clustering| Partitional clustering]).
  
 
ILMSImage implements two methods of cluster analysis:  
 
ILMSImage implements two methods of cluster analysis:  
  
# A '''k-Means algorithm''' which represents a frequently used method for finding structures in large amounts of data. The number of clusters of the clusters which should be found is determined beforehand.  After a random initialization every cell is assigned to the cluster the central feature vector of which is most similar to that of the cell. If this initial assignment is completed, the cluster centers are recalculated and every cell is compared to existing clusters again. These steps are repeated until none of the assignments change again - the the cluster analysis is completed.
+
# A '''k-Means algorithm''' which represents a frequently used method for finding structures in large amounts of data. The number of clusters of the clusters which should be found is determined beforehand.  After a random initialization every cell is assigned to the cluster the central feature vector of which is most similar to that of the cell. If this initial assignment is completed, the cluster centers are recalculated and every cell is compared to existing clusters again. These steps are repeated until none of the assignments change again - the cluster analysis is completed.
# Another method for partitional clutering is the algorithm of '''[http://en.wikipedia.org/wiki/Self-organizing_map| self-organizing maps]''' which was originally developed by Teuvu Kohonen. It is an artificial neuronal network which is able to project a multi-dimensional feature space functioning as a data source on a two-dimensional range of values. This characteristic explains the relation to the ''map''. The detailed funcitoning of this algorithm is described at another point.
+
# Another method for partitional clustering is the algorithm of '''[http://en.wikipedia.org/wiki/Self-organizing_map| self-organizing maps]''' which was originally developed by Teuvu Kohonen. It is an artificial neuronal network which is able to project a multi-dimensional feature space functioning as a data source on a two-dimensional range of values. This characteristic explains the relation to the ''map''. The detailed functioning of this algorithm is described at another point.
  
Die von ILMSImage erzeugten Zellen und ihre abgeleiteten Eigenschaften stellen ebenfalls einen mehrdimensionalen Merkmalsraum dar. Soll dieser nun unüberwacht klassifiziert, d.h. geclustert werden, so muss diese Option im entsprechenden Teil des ILMSImage-Klassifikationspanels aktiviert werden. Über ''Type'' lässt sich eine der beiden oben genannten Methoden auswählen, der Parameter ''Tolerance'' (Wertebereich zwischen 0,01 und 0,50) steuert die Sensitivität des Prozesses. Ein höherer Wert entspricht dabei einer einer höheren erlaubten Merkmalsvariabilität innerhalb der zu erzeugenden Cluster - ihre Anzahl nimmt dann also ab. Ein geringer Wert für die Toleranz erzeugt eine höhere Zahl von Klassen, da die erlaubte Variabilität der Merkmale innerhalb dieser dann geringer ist.
+
The cells created by ILMSImage and their derived features also represent a multi-dimensional feature space. If this is supposed to be classified in an unsupervised manner, i.e. clustering, this option has to activated in the corresponding part of the ILMSImage classification panel. By ''Type'' one of the two above-mentioned method can be selected, the parameter ''tolerance'' (range of values from 0.01 to 0.50) controls the sensitivity of the processes. A higher value corresponds to a higher acceptable variability of features within the  clusters which are to be generated - hence their number decreases. A low value of tolerance generates a high number of classes since the acceptable variability of features within those is lower.
  
Die andauernde Prozessierung wird durch ein entsprechendes Fenster angezeigt. Je nach Anzahl der ausgewählten Attribute und der eingestellten Sensitivität kann die Berechnungsdauer stark variieren. Nach erfolgreichem Abschluss des Prozesses wird der Benutzer über die Anzahl der erzeugten Zelltypenklassen informiert:
+
The continuing processing is indicated by a corresponding window. According to the number of selected attributes and to the selected sensitivity the duration of calculation time can vary greatly. After successful completion of the process the user is informed about the number of generated cell type classes:
  
[[Bild:ilms_img_qgis_work_30.png|Benachrichtigungsfenster nach Abschluss der unüberwachten Klassifikation]]
+
[[File:ilms_img_qgis_work_30.png|Benachrichtigungsfenster nach Abschluss der unüberwachten Klassifikation]]
  
Bleibt die Auswahlbox ''Unsupervised classification'' aktiviert, kann mit einem Klick auf ''Visualize results'' das Ergebnis der unüberwachten Klassifikation in die aktuelle Kartenansicht geladen werden. Die zur Darstellung verwendeten Farben sind Zufallsfarben, das heißt sie besitzen keinerlei semantische Aussagekraft über die thematische Klasse einer Zelle. Sie verdeutlichen lediglich ihre gemeinsame Zugehörigkeit zu einer Zelltypenklasse. In der Legende erscheint der Rasterdatenlayer in der Gruppe ''ILMSImage Exports'' unter der Bezeichnung <code><projektname>_sample</code>.
+
If the check box ''Unsupervised classification'' stays activated, one click on ''Visualize results'' can load the result of the unsupervised classification into the current map view. The colors used for the visualization are random colors, this is they do not possess any semantic significance in relation to the thematic class of a cell. They only illustrate their common belonging to one cell type class. In the key the raster data layer appears in the group ''ILMSImage Exports'' specified as <code><project name>_sample</code>.
  
[[Bild:ilms_img_qgis_work_31.png|800px|Im aktuellen QuantumGIS-Projekt visualisiertes Ergebnis der unüberwachten Klassifkation]]
+
[[File:ilms_img_qgis_work_31.png|800px|Im aktuellen QuantumGIS-Projekt visualisiertes Ergebnis der unüberwachten Klassifkation]]
  
Es ist zu beachten, dass immer nur ein Ergebnis dieses Arbeitsschritts existiert, das für die weitere thematischen Klassifikation verwendet wird. Wurde ein solches Ergebnis nach seiner Erzeugung nicht wie oben beschrieben visualisiert, so entspricht der in der Kartenansicht angezeigte Rasterlayer mit hoher Wahrscheinlichkeit nicht dieser Datengrundlage.
+
It should be noted that only one result can exist in this task which can be used for further thematic classification. If such a result has not been visualized after its generation as described above, the raster layer shown in the map view most likely does not correspond to this data basis.
  
===Definition von Referenzgebieten===
+
===Defining Reference Areas===
  
Referenz- oder Trainingsgebiete erlauben es dem Benutzer, das für sein Arbeitsziel zweckmäßige Klassifikationsschema zu definieren und an ILMSImage zu übermitteln. Die abgegrenzten Gebiete und deren Attribute werden dabei aus einem Vektordatenlayer, speziell einer Shapedatei, übernommen. Die Attributtabelle dieser Datei muss - um einen korrekten Import zu gewährleisten - einem bestimmten Format entsprechen, das in der Folge beschrieben ist.  
+
Reference and training areas can help the user to define the appropriate classification scheme for his/her objective and convey it to ILMSImage. The separated areas and their attributes are taken from the vector data layer, particularly a shapefile. The attribute table of this file has to correspond to a certain format - to allow correct import - which is described below.
  
 
{| border="1" cellspacing="0" cellpadding="10" style="border-collapse: collapse;"
 
{| border="1" cellspacing="0" cellpadding="10" style="border-collapse: collapse;"
 
|-
 
|-
! width="20%" | Feldname !! width="20%" | Felddatentyp !! width="60%" | Beschreibung
+
! width="20%" | Field name!! width="20%" | Field data type!! width="60%" | Description
 
|-
 
|-
| <code><b>thema</b></code> || Ganzzahl / Integer || Nummer zur eindeutigen Identifizierung der verschiedenen Referenzklassen. Dabei ist es wichtig, dass die '''Zählung mit eins beginnt''' und '''fortlaufend''' ist.
+
| <code><b>thema</b></code> || Integer || Number for clear identification of the different reference classes. It is important that the '''counting begins with one''' and is '''consecutive'''.
 
|-
 
|-
| <code><b>desc</b></code> || Text / String || Verbale Beschreibung der Referenzklasse. Im Grunde handelt es sich bei diesem Feld um eine optionale Angabe, es kann jedoch als gute Praxis angesehen werden, den sonst nur mit Nummer bestimmten Referenzklassen auch einen Namen zu geben.
+
| <code><b>desc</b></code> || String || Verbal description of the reference class. It is basically a field which is optional information but can be seen as good practice to give names to reference classes which are otherwise only defined by numbers.  
 
|}
 
|}
  
Die Shapedatei kann weitere Felder besitzen - um sie zu einem gültigen Referenzdatensatz zu machen genügt es, dass die beiden oben genannten Felder in der Attributtabelle zu finden sind. Auf diese Art und Weise kann grundsätzlich jede Flächenshapedatei - mit den entsprechenden Anpassungen - als Referenz für ILMSImage verwendet werden.
+
The shapefile can have additional fields - to create a valid reference dataset it is sufficient that both of the above-mentioned fields can be found in the attribute table. In this way basically every areal shapefile - including necessary modifications - can be used as reference for ILMSImage.
  
Um dem Nutzer die Erstellung eines neuen Referenzdatensatzes zu erleichtern, stellt das QuantumGIS-Plugin optional eine passende Shapedatei mit der richtigen Struktur bereit, die im weiteren Verlauf mit den Gebietsgrenzen und deren Attributierung vervollständigt werden kann. Um diese Funktion zu nutzen genügt - nach erstmaliger Aktivierung des gesamten Abschnitts ''Reference data import'' mittels der entsprechenden Auswahlbox - ein Klick auf ''Create reference layer template''. Der folgende Dialog erfragt den Namen der zu erstellenden Datei, standardmäßig ist die Kombination <code><projektname>_ref</code> voreingestellt und bittet den Nutzer die voraussichtliche Anzahl der thematischen Klassen festzulegen. Diese Angabe ist nicht endgültig, es können im Nachhinein noch Klassen hinzugefügt oder weggelassen werden - sie sorgt allein dafür, dass für die Definition der Referenzgebiete bereits die festgelegte Anzahl von Klassen (mit unterschiedlich farblicher Darstellung) bereitgestellt wird.
+
In order to facilitate the generation of a new reference data set for the user, the QuantumGIS plug-in makes a corresponding shapefile available which has the right structure and can be complemented with area borders and their attributions. The function can be activated by clicking on ''Create reference layer template'', after the entire section ''Reference data import'' has been activated by using the corresponding check box. The following dialog requests the name of the file which is supposed to be generated (the default setting is <code><project name>_ref</code>) and asks the user to determine the estimated number of thematic classes. This indication is not final, subsequently classes can be added or removed - the only function is to make the determined number of classes (with different visual representation in terms of color) available for the definition of reference areas.
  
[[Bild:ilms_img_qgis_work_32.png|Auswahldialog zur Erzeugung eines neuen Referenzdatensatzes mit voraussichtlicher Klassenanzahl]]
+
[[File:ilms_img_qgis_work_32.png|Auswahldialog zur Erzeugung eines neuen Referenzdatensatzes mit voraussichtlicher Klassenanzahl]]
  
Für den Fall, dass der zu erzeugenden Referenzdatensatz bereits existiert, wird der der Nutzer darüber informiert und um eine entsprechende Entscheidung gebeten:
+
In case the reference data set which should be generated already exists, the user will be informed and asked to make a decision:
  
[[Bild:ilms_img_qgis_work_33.png|Benachrichtigung im Falle, dass der zu erzeugende Referenzdatensatz bereits existiert]]
+
[[File:ilms_img_qgis_work_33.png|Benachrichtigung im Falle, dass der zu erzeugende Referenzdatensatz bereits existiert]]
  
Wurde der Layer erzeugt, erscheint er in der Ebenenübersicht des aktuellen QuantumGIS-Projekts. Dort kann er auch mittels eines Rechtsklicks und der Auswahl von ''Bearbeitungsmodus umschalten'' in einen editierfähigen Zustand versetzt werden. Das ist notwendig, um in der Folge die Geometrien der Trainingsgebiete und ihre Attributierung in den Referenzdatensatz einzutragen.
+
When the layer has been generated, it appears in the level overview of the current QuantumGIS project. There it can also be transferred into an editable mode by right-clicking on and selecting ''Bearbeitungsmodus umschalten'' (change editing mode). This is necessary to register the geometries of the training areas and their attribution into the reference dataset.
  
[[Bild:ilms_img_qgis_work_34.png|Umschalten in den Bearbeitungsmodus]]
+
[[File:ilms_img_qgis_work_34.png|Umschalten in den Bearbeitungsmodus]]
  
Nun lassen sich mittels des Werkzeugs ''Polygon digitaliseren'' neue Referenzgebiete erzeugen. Dies geschieht i.d.R. am besten im Rückgriff auf die originalen Bilddaten, je nach Arbeitsziel und Klassifikationsschema können jedoch auch andere Visualisierungen gute Anhaltspunkte für die Trainingsgebietsabgrenzung liefern. In der folgenden Darstellung ist an einem Detail die Erzeugung eines Referenzgebiets für die Klasse ''Settlement/Siedlung'' illustriert.
+
Now the tool ''Polygon digitalisieren'' (digitize polygon) can be used to generate new reference areas. This usually happens when re-accessing original image data; according to the objective and classification scheme also other visualizations can deliver good indications for borders of training areas. The following image illustrates a detail of the generation of a reference area for the class ''Settlement/Siedlung''.  
  
 
<ol>
 
<ol>
<li>Zunächst werden die Stützpunkte des Polygons durch Linksklicks an den entsprechenden Koordinaten erzeugt. Es ist dabei nicht unbedingt notwendig, sehr genau vorzugehen, ILMSImage ist in der Lage, eventuelle Randbereiche durch statistische Kriterien von den Kernarealen eines Referenzgebiets zu trennen. Statt Genauigkeit ist eine Vergrößerung der Stichprobe - durch Erstellung vieler Referenzgebiete für eine thematische Klasse - oft sinnvoller. <br/><br/>
+
<li>First the support points of the polygon are generated by left-clicking at the corresponding coordinates. It is not necessary to be very precise since ILMSImage is able to separate potential peripheries from central areas of a reference area by statistic criteria. Instead of being accurate it often makes more sense to increase the sample by generating various reference areas for a thematic class. <br/><br/>
[[Bild:ilms_img_qgis_work_35.png|400px|Digitalisieren eines Referenzpolygons]]<br/><br/></li>
+
[[File:ilms_img_qgis_work_35.png|400px|Digitalisieren eines Referenzpolygons]]<br/><br/></li>
<li>Sind alle Stützpunkte gesetzt, wird durch einen Rechtsklick die Digitalisierung des Polygons abgeschlossen. In der folgenden Eingabemaske erfolgt die Angabe der Klassen-ID sowie der entsprechenden Beschreibung. Es sei daran erinnert, dass die Nummerierung der Klassen bei eins beginnen und fortlaufend sein muss.<br/><br/>
+
<li>When all support points are set, a right-click finishes the digitization of the polygon. The following input mask the class ID and the corresponding description are indicated. It should be noted again that the numbering of classes starts from one and has to be consecutive.<br/><br/>
[[Bild:ilms_img_qgis_work_36.png|400px|Maske zur Eingabe der Attribute]]<br/><br/></li>  
+
[[File:ilms_img_qgis_work_36.png|400px|Maske zur Eingabe der Attribute]]<br/><br/></li>  
<li>Nach Bestätigung der Eingabemaske wird das Polygon in der für diese ID zufällig erzeugten Farbe angezeigt. <br/><br/>
+
<li>After confirming the input mask, the polygon is shown in the color which is randomly selected for this ID.<br/><br/>
[[Bild:ilms_img_qgis_work_37.png|400px|Vollständiges Referenzgebiet in Zufallsfarbendarstellung]]<br/><br/></li>
+
[[File:ilms_img_qgis_work_37.png|400px|Vollständiges Referenzgebiet in Zufallsfarbendarstellung]]<br/><br/></li>
<li>Es ist hilfreich, die Farbe der Darstellung dem dargestellten Sachverhalt so anzupassen, dass entweder Autoplausibilität entsteht (Wald = grün, Wasser = blau) oder auf gängige Anwendung bestimmter Farben (Siedlung = rötlich, Landwirtschaft = erdfarben) zurückgegriffen wird. Durch Verwendung der oben beschriebenen Vorlage muss dazu nicht mehr die ganze Darstellung des Layers, sondern nur noch die Farbe der einzelnen thematischen Klasse angepasst werden. Das dazu benötigte Fenster ist durch einen Linksklick auf die zu ändernde Farbe erreichbar. Dort kann auch die Beschriftung (von ''reference class 1'' auf ''Settlement'') aktualisiert werden. <br/><br/>
+
<li>It is useful to modify the color of the representation according to the facts, so that either auto-plausibility is established (forest = green, water = blue) or the common usage of certain colors (settlement = red, agriculture = earth-colored) is applied. Using the above-described template means that no longer the entire representation of the layer has to be adjusted but only the color of individual thematic classes. The required window can be accessed by left-clicking on the color which is to be modified. There also the designation can be updated (from ''reference class 1'' to ''Settlement''). <br/><br/>
[[Bild:ilms_img_qgis_work_38.png|400px|Fenster zur Änderung von Darstellungsfarbe und Beschreibung]]<br/><br/></li>
+
[[File:ilms_img_qgis_work_38.png|400px|Fenster zur Änderung von Darstellungsfarbe und Beschreibung]]<br/><br/></li>
<li>Das Ergebnis ist ein erstes Referenzgebiet einer bestimmten thematischen Klasse, im illustrierten Fall für die Klasse ''Siedlung''.<br/><br/>
+
<li>The results is a first reference areas of a specific thematic class, the illustrated example shows the class ''Settlement/Siedlung''.<br/><br/>
[[Bild:ilms_img_qgis_work_39.png|400px|Vollständiges Referenzgebiet in dem Sachverhalt angepasster Darstellung]]<br/><br/></li>
+
[[File:ilms_img_qgis_work_39.png|400px|Vollständiges Referenzgebiet in dem Sachverhalt angepasster Darstellung]]<br/><br/></li>
 
</ol>
 
</ol>
  
Das beschriebene Vorgehen wird wiederholt, bis ausreichend Referenzgebiete definiert wurden. Das wiederholten Umschalten und die Bestätigung des folgenden Dialogfensters sichern den Referenzdatensatz. Werden auch die Farben und die Beschriftungen der Klassen angepasst, ergibt sich dann für das Beispiel der Rot u.U. folgendes Bild:
+
The above-described procedure is repeated until a sufficient number of reference areas has been defined. A repeated switching and confirming the following dialog window back up the reference dataset. If the colors and designations of the classes are modified as well, the example of the Rot shows the following image:
  
[[Bild:ilms_img_qgis_work_40.png|800px|Vollständig definierter Referenzdatenlayer überlagert über der Grenze des Untersuchungsgebiets]]
+
[[File:ilms_img_qgis_work_40.png|800px|Vollständig definierter Referenzdatenlayer überlagert über der Grenze des Untersuchungsgebiets]]
  
===Import von Referenzgebieten===
+
===Importing Reference Areas===
  
Um im Zuge der thematischen Klassifikation verwendet werden zu können, muss der definierte Referenzdatensatz in das ILMSImage-Projekt importiert werden. Er muss dazu zunächst einmal im aktuellen QuantumGIS-Projekt geladen sein - egal auf welchem Wege er erstellt wurde. In der Auswahlliste ''Reference data layer'' sind alle geladenen Vektorebenen aufgelistet, welche die oben beschriebene Struktur besitzen und damit von ILMSImage als importierbar betrachtet werden. Nach dem der gewünschte Layer selektiert wurde, startet ein Klick auf ''Conduct classification'' den eigentlichen Referenzdatenimport. Dabei wird der Vektordatensatz ins Rasterformat umgewandelt und der Attributliste der Zellen hinzugefügt. Das sollte i.d.R. sehr schnell passieren. Das fertige Ergebnis lässt sich durch einen Klick auf ''Visualize results'' in der aktuellen Kartenansicht darstellen. Für den oben beispielhaft gezeigten Referenzdatensatz könnte das Ergebnis - abhängig von den zu Grunde liegenden Zellgeometrien - so aussehen:
+
In order to make the defined reference dataset suitable for thematic classification, it has to be imported into the ILMSImage project. This means, first of all, that the data have to be loaded in the current QuantumGIS project - no matter how it has been generated. In the list box ''Reference data layer'' all loaded vector levels are listed which have the above-described structure and are therefore considered importable by ILMSImage. After the layer required has been selected, clicking on ''Conduct classification'' starts the actual import of reference data. This transforms the vector dataset into raster format and added to the attribute list of the cells. This is supposed to happen usually very fast. The complete result can be represented in the current map view by clicking on ''Visualize results''. The result for the above-described example of a reference dataset can be like this (subject to the underlying cell geometries):
  
[[Bild:ilms_img_qgis_work_41.png|800px|Importierter und visualisierter Referenzdatensatz]]
+
[[File:ilms_img_qgis_work_41.png|800px|Importierter und visualisierter Referenzdatensatz]]
  
Die für die Visualisierung verwendeten Farben übernimmt der Prozess direkt aus der Darstellung des entsprechenden Vektordatenlayers. Um die Themenklassen auseinander halten zu können, ist es daher empfehlenswert, schon bei der Definition unterschiedliche Farben für sie zu verwenden. Wenn diese dabei autoplausibel bzw. entsprechend gängiger Assoziationen folgend (siehe oben) eingesetzt werden, ist die rasche Erkennbarkeit des Referenzdatensatzes im Rasterformat - und später des eigentlichen Klassifikationsergebnisses - gewährleistet.
+
The process applies the colors used for visualization directly from the representation of the corresponding vector data layer. For a better differentiation of the thematic classes it is recommended to use different colors for them when they are defined. If they are auto-plausible or correspond to common associations (see above), a quick recognition of the reference dataset in raster format - and a subsequent result of the actual classification  - is guaranteed.
  
===Endgültige Klassifikation===
+
===Final Classification===
  
Die Auswahlbox ''Supervised classification'' aktiviert den letzten Arbeitsschritt im Rahmen des ILMSImage-Panels für die thematische Klassifikation. Dazu werden die Ergebnisse der Clusteranalyse und des Referenzdatenimports miteinander verknüpft.
+
The check box ''Supervised classification'' activates the last task in the context of the ILMSImage panel for thematic classification. For this purpose the results of the cluster analysis and of the reference data import are linked to one another.
  
 +
The final result of the classification is available in raster format as a GeoTIFF file to allow maximum comparability.
  
Das endgültige Klassifikationsergebnis liegt im Rasterformat als GeoTIFF-Datei vor, um größtmögliche Kompatibilität zu gewährleisten.
+
[[File:ilms_img_qgis_work_42.png|800px|Endgültige thematische Klassifikation des Untersuchungsgebiets]]
 
+
[[Bild:ilms_img_qgis_work_42.png|800px|Endgültige thematische Klassifikation des Untersuchungsgebiets]]
+

Latest revision as of 00:52, 13 February 2012

Contents

ILMSImage Classification

Introduction

ILMSImage Classification is part of the ILMSImage plug-in for QuantumGIS and in this context it carries out a two-step thematic analysis, comprising unsupervised and supervised classification.

Like various ILMSImage panels it consists of two components, the ILMSImage Project Information in the upper section and the actual tools in the lower section.

Das ILMSImage-Panel zur thematischen Klassifikation

Thematic Classification

Background

The process of thematic classification within ILMSImage is composed of three sub-tasks, (1) an unsupervised classification of the existing cells and their features as cell type classes, (2) the definition of reference or training areas using a vector data layer and (3) the actual thematic classification which is based on results of both preceding tasks. Below every step is presented and described in detail.

Unsupervised Classification into Cell Type Classes

The unsupervised classification refers to generated geometries during the cell creation and the calculated attributes for the geometries to derive cell type classes from both components. They are group of cells which - no matter which geographical location in the image or location to one another - have similar features. Simply put, in this task e.g. elongated and dark cells are separated from those which are rather round and bright - only that the corresponding decision in reality is not limited to the named features but is based on all features generated during attribute calculation. The basic concept of this task therefore corresponds to the partitional cluster analysis (Partitional clustering).

ILMSImage implements two methods of cluster analysis:

  1. A k-Means algorithm which represents a frequently used method for finding structures in large amounts of data. The number of clusters of the clusters which should be found is determined beforehand. After a random initialization every cell is assigned to the cluster the central feature vector of which is most similar to that of the cell. If this initial assignment is completed, the cluster centers are recalculated and every cell is compared to existing clusters again. These steps are repeated until none of the assignments change again - the cluster analysis is completed.
  2. Another method for partitional clustering is the algorithm of self-organizing maps which was originally developed by Teuvu Kohonen. It is an artificial neuronal network which is able to project a multi-dimensional feature space functioning as a data source on a two-dimensional range of values. This characteristic explains the relation to the map. The detailed functioning of this algorithm is described at another point.

The cells created by ILMSImage and their derived features also represent a multi-dimensional feature space. If this is supposed to be classified in an unsupervised manner, i.e. clustering, this option has to activated in the corresponding part of the ILMSImage classification panel. By Type one of the two above-mentioned method can be selected, the parameter tolerance (range of values from 0.01 to 0.50) controls the sensitivity of the processes. A higher value corresponds to a higher acceptable variability of features within the clusters which are to be generated - hence their number decreases. A low value of tolerance generates a high number of classes since the acceptable variability of features within those is lower.

The continuing processing is indicated by a corresponding window. According to the number of selected attributes and to the selected sensitivity the duration of calculation time can vary greatly. After successful completion of the process the user is informed about the number of generated cell type classes:

Benachrichtigungsfenster nach Abschluss der unüberwachten Klassifikation

If the check box Unsupervised classification stays activated, one click on Visualize results can load the result of the unsupervised classification into the current map view. The colors used for the visualization are random colors, this is they do not possess any semantic significance in relation to the thematic class of a cell. They only illustrate their common belonging to one cell type class. In the key the raster data layer appears in the group ILMSImage Exports specified as <project name>_sample.

Im aktuellen QuantumGIS-Projekt visualisiertes Ergebnis der unüberwachten Klassifkation

It should be noted that only one result can exist in this task which can be used for further thematic classification. If such a result has not been visualized after its generation as described above, the raster layer shown in the map view most likely does not correspond to this data basis.

Defining Reference Areas

Reference and training areas can help the user to define the appropriate classification scheme for his/her objective and convey it to ILMSImage. The separated areas and their attributes are taken from the vector data layer, particularly a shapefile. The attribute table of this file has to correspond to a certain format - to allow correct import - which is described below.

Field name Field data type Description
thema Integer Number for clear identification of the different reference classes. It is important that the counting begins with one and is consecutive.
desc String Verbal description of the reference class. It is basically a field which is optional information but can be seen as good practice to give names to reference classes which are otherwise only defined by numbers.

The shapefile can have additional fields - to create a valid reference dataset it is sufficient that both of the above-mentioned fields can be found in the attribute table. In this way basically every areal shapefile - including necessary modifications - can be used as reference for ILMSImage.

In order to facilitate the generation of a new reference data set for the user, the QuantumGIS plug-in makes a corresponding shapefile available which has the right structure and can be complemented with area borders and their attributions. The function can be activated by clicking on Create reference layer template, after the entire section Reference data import has been activated by using the corresponding check box. The following dialog requests the name of the file which is supposed to be generated (the default setting is <project name>_ref) and asks the user to determine the estimated number of thematic classes. This indication is not final, subsequently classes can be added or removed - the only function is to make the determined number of classes (with different visual representation in terms of color) available for the definition of reference areas.

Auswahldialog zur Erzeugung eines neuen Referenzdatensatzes mit voraussichtlicher Klassenanzahl

In case the reference data set which should be generated already exists, the user will be informed and asked to make a decision:

Benachrichtigung im Falle, dass der zu erzeugende Referenzdatensatz bereits existiert

When the layer has been generated, it appears in the level overview of the current QuantumGIS project. There it can also be transferred into an editable mode by right-clicking on and selecting Bearbeitungsmodus umschalten (change editing mode). This is necessary to register the geometries of the training areas and their attribution into the reference dataset.

Umschalten in den Bearbeitungsmodus

Now the tool Polygon digitalisieren (digitize polygon) can be used to generate new reference areas. This usually happens when re-accessing original image data; according to the objective and classification scheme also other visualizations can deliver good indications for borders of training areas. The following image illustrates a detail of the generation of a reference area for the class Settlement/Siedlung.

  1. First the support points of the polygon are generated by left-clicking at the corresponding coordinates. It is not necessary to be very precise since ILMSImage is able to separate potential peripheries from central areas of a reference area by statistic criteria. Instead of being accurate it often makes more sense to increase the sample by generating various reference areas for a thematic class.

    Digitalisieren eines Referenzpolygons

  2. When all support points are set, a right-click finishes the digitization of the polygon. The following input mask the class ID and the corresponding description are indicated. It should be noted again that the numbering of classes starts from one and has to be consecutive.

    Maske zur Eingabe der Attribute

  3. After confirming the input mask, the polygon is shown in the color which is randomly selected for this ID.

    Vollständiges Referenzgebiet in Zufallsfarbendarstellung

  4. It is useful to modify the color of the representation according to the facts, so that either auto-plausibility is established (forest = green, water = blue) or the common usage of certain colors (settlement = red, agriculture = earth-colored) is applied. Using the above-described template means that no longer the entire representation of the layer has to be adjusted but only the color of individual thematic classes. The required window can be accessed by left-clicking on the color which is to be modified. There also the designation can be updated (from reference class 1 to Settlement).

    Fenster zur Änderung von Darstellungsfarbe und Beschreibung

  5. The results is a first reference areas of a specific thematic class, the illustrated example shows the class Settlement/Siedlung.

    Vollständiges Referenzgebiet in dem Sachverhalt angepasster Darstellung

The above-described procedure is repeated until a sufficient number of reference areas has been defined. A repeated switching and confirming the following dialog window back up the reference dataset. If the colors and designations of the classes are modified as well, the example of the Rot shows the following image:

Vollständig definierter Referenzdatenlayer überlagert über der Grenze des Untersuchungsgebiets

Importing Reference Areas

In order to make the defined reference dataset suitable for thematic classification, it has to be imported into the ILMSImage project. This means, first of all, that the data have to be loaded in the current QuantumGIS project - no matter how it has been generated. In the list box Reference data layer all loaded vector levels are listed which have the above-described structure and are therefore considered importable by ILMSImage. After the layer required has been selected, clicking on Conduct classification starts the actual import of reference data. This transforms the vector dataset into raster format and added to the attribute list of the cells. This is supposed to happen usually very fast. The complete result can be represented in the current map view by clicking on Visualize results. The result for the above-described example of a reference dataset can be like this (subject to the underlying cell geometries):

Importierter und visualisierter Referenzdatensatz

The process applies the colors used for visualization directly from the representation of the corresponding vector data layer. For a better differentiation of the thematic classes it is recommended to use different colors for them when they are defined. If they are auto-plausible or correspond to common associations (see above), a quick recognition of the reference dataset in raster format - and a subsequent result of the actual classification - is guaranteed.

Final Classification

The check box Supervised classification activates the last task in the context of the ILMSImage panel for thematic classification. For this purpose the results of the cluster analysis and of the reference data import are linked to one another.

The final result of the classification is available in raster format as a GeoTIFF file to allow maximum comparability.

Endgültige thematische Klassifikation des Untersuchungsgebiets

Personal tools
In other languages