Supervised classification in QGIS - Revision history

Smitchs: /* Set Region of Interests (ROI) */

2025-10-26T18:06:26Z

Set Region of Interests (ROI)

Oliver McMillan: /* Obtaining the Data */

2025-10-01T01:07:29Z

Obtaining the Data

Oliver McMillan: /* Obtaining the Data */

2025-10-01T01:06:56Z

Obtaining the Data

Oliver McMillan: /* Calculate the Kappa Coefficient */

2025-09-30T14:54:02Z

Calculate the Kappa Coefficient

Oliver McMillan at 14:53, 30 September 2025

2025-09-30T14:53:27Z

Benjamincolbourne: /* Set Region of Interests (ROI) */

2021-10-02T17:46:14Z

Set Region of Interests (ROI)

Benjamincolbourne: /* Set Region of Interests (ROI) */

2021-10-02T17:31:09Z

Set Region of Interests (ROI)

Benjamincolbourne: /* Clip the Data */

2021-10-02T17:09:57Z

Clip the Data

Benjamincolbourne: /* Assessing the classification */

2021-10-02T16:59:03Z

Assessing the classification

Benjamincolbourne: /* Calculate the Kappa Coefficient */

2021-10-02T16:58:49Z

Calculate the Kappa Coefficient

@@ Line 86: / Line 86: @@
 We can now begin with the supervised classification. Make sure you see the '''SCP & Dock''' at your surface. If not, clicking this button [[File:OpenSCP.JPG]] in the toolbar will open it.
-===Set Region of Interests (ROI)===
+===Set Region of Interest (ROI)===
 First, you must create a file where the ROIs can be saved. To do so, click '''Training input''' then this button:

← Older revision		Revision as of 01:07, 1 October 2025
Line 20:		Line 20:


	'''Copernicus Browser''' is the European Space Agency's equivalent to USGS earth explorer but focuses on Sentinel data. Go to [https://browser.dataspace.copernicus.eu/?zoom=5&lat=50.16282&lng=20.78613&themeId=DEFAULT-THEME&visualizationUrl=U2FsdGVkX188tgNdFi%2FL3tpFa0q6CLBfOt2NICr3ZIWELzc2vM9jWiPOR2XjZFetLoOWjtVb7GQWUj9DGUK2Ek1wE%2BpqvwMHQ7p1eaFithPVU1D2o5lP73lP%2BSRcZwDY&datasetId=S2_L2A_CDAS&demSource3D=%22MAPZEN%22&cloudCoverage=30&dateMode=SINGLE Copernicusbrowser.com] and create an account if you don't already have one.		'''Copernicus Browser''' is another option, it is the European Space Agency's equivalent to USGS earth explorer but focuses on Sentinel data. Go to [https://browser.dataspace.copernicus.eu/?zoom=5&lat=50.16282&lng=20.78613&themeId=DEFAULT-THEME&visualizationUrl=U2FsdGVkX188tgNdFi%2FL3tpFa0q6CLBfOt2NICr3ZIWELzc2vM9jWiPOR2XjZFetLoOWjtVb7GQWUj9DGUK2Ek1wE%2BpqvwMHQ7p1eaFithPVU1D2o5lP73lP%2BSRcZwDY&datasetId=S2_L2A_CDAS&demSource3D=%22MAPZEN%22&cloudCoverage=30&dateMode=SINGLE Copernicusbrowser.com] and create an account if you don't already have one.

	For a tutorial on how to use the browser and get the imagery you're interested in click the little "i" at the top right of the page.		For a tutorial on how to use the browser and get the imagery you're interested in click the little "i" at the top right of the page.

@@ Line 18: / Line 18: @@
 It is always easier to work with cloud-free pictures, otherwise, you have to use a cloud mask.
+'''Copernicus Browser''' is the European Space Agency's equivalent to USGS earth explorer but focuses on Sentinel data. Go to [https://browser.dataspace.copernicus.eu/?zoom=5&lat=50.16282&lng=20.78613&themeId=DEFAULT-THEME&visualizationUrl=U2FsdGVkX188tgNdFi%2FL3tpFa0q6CLBfOt2NICr3ZIWELzc2vM9jWiPOR2XjZFetLoOWjtVb7GQWUj9DGUK2Ek1wE%2BpqvwMHQ7p1eaFithPVU1D2o5lP73lP%2BSRcZwDY&datasetId=S2_L2A_CDAS&demSource3D=%22MAPZEN%22&cloudCoverage=30&dateMode=SINGLE Copernicusbrowser.com] and create an account if you don't already have one.
+For a tutorial on how to use the browser and get the imagery you're interested in click the little "i" at the top right of the page.
+[[File:Copernicus Browser information icon.png|500px|Click the little 'i' in the top right of the browser.]]
+Images can be found by filtering by date, aoi, and product type. You can search for the same image used above by copying and pasting the image id into the search bar.
+[[File:Copernicus Browser Search.png|300px|]]
 ===Unpack the Data===

@@ Line 167: / Line 167: @@
 The SCP provides a lot of options to achieve a good classification result. It depends on the approach, how much time one wants to spend to improve the classification. The SCP provides even more options to improve the ROIs while altering the spectral signatures for different classes. Nonetheless, it will not be possible to classify every single pixel right.
-===Calculate the Kappa Coefficient===
+===Calculate the Users and Producers Accuracy===
 A quantitative method to assess the classification is to calculate Users and Producers accuracy. First, you have to create a new layer with ROIs and set again ROIs for the four classes to have a reference ground. You cannot use the ROIs you used for the classification because you want to compare the classification with undependable training input. After you created various ROIs open the SCP and go to '''Postprocessing''', '''Accuracy'''. As your input layer choose your best classification result. The reference raster layer will be the new ROIs you just set:

@@ Line 168: / Line 168: @@
 ===Calculate the Kappa Coefficient===
-A quantitative method to assess the classification is to calculate the Kappa Coefficient. First, you have to create a new layer with ROIs and set again ROIs for the four classes to have a reference ground. You cannot use the ROIs you used for the classification because you want to compare the classification with undependable training input. After you created various ROIs open the SCP and go to '''Postprocessing''', '''Accuracy'''. As your input layer choose your best classification result. The reference raster layer will be the new ROIs you just set:
+A quantitative method to assess the classification is to calculate Users and Producers accuracy. First, you have to create a new layer with ROIs and set again ROIs for the four classes to have a reference ground. You cannot use the ROIs you used for the classification because you want to compare the classification with undependable training input. After you created various ROIs open the SCP and go to '''Postprocessing''', '''Accuracy'''. As your input layer choose your best classification result. The reference raster layer will be the new ROIs you just set:
 [[File:Accuracy assessment.JPG|500px]]
-The output will tell you the accuracy for each class and the overall accuracy. The Kappa scale is from 0 to 1, 0 means the classification is not better than random, 1 means the classification is highly accurate.
+The output will tell you the accuracy for each class and the overall accuracy.
 [[File:ErrorMatrix.JPG|650px|thumb|none|Accuracy assessment Report of the Minimum Distance classification]]
@@ Line 181: / Line 179: @@
+The output will tell you the accuracy for each class and the overall accuracy. Users accuracy is a measure of false positives or type 1 errors. Producers accuracy is a measure of false negatives or type 2 errors. For both users and producers accuracy higher percentages mean lower errors rates. They are two sides of the same coin and give us an idea of the types of inaccuracies that we have in our classification. We can determine which classes were confused for others in the error matrix. You can learn how to read the error matrix and learn more on user and producer accuracies [https://pro.arcgis.com/en/pro-app/latest/help/analysis/image-analyst/accuracy-assessment.htm here]. The overall accuracy is a combination of both measures.
-In the first picture you see the assessment report of the Minimum Distance algorithm and on the second the one from the Spectral Angle Mapping. Comparing both, the overall Kappa Coefficient of the Spectral Angle Mapping is a bit higher (0.943) than the one of the Maximum Distance (~0.913). However, both overall Kappa Coefficients values are very high. This is questionable and probably because too little ROIs were set in the second ROI ground reference Layer.
 ==Conclusion==

@@ Line 87: / Line 87: @@
 [[File:Create_ROI.JPG]]
+Note: '''Save''' often when creating ROI's. This is a time consuming process and if the program were to crash you could loose many minutes of your time.
 In the following picture, the first ROI is in the lake. You can see that the macro class ('''MC ID''') is named ''Water'' and the subclass ('''C ID''') ''Lake''. Make sure to change the '''MC ID''' name when creating ROI's for different land classes. In this tutorial, only the macro classes will be significant, since it is a basic classification with only five different classes. If you want to have more specific classes you can use the subclasses. Save the ROI.

← Older revision		Revision as of 17:31, 2 October 2021
Line 88:		Line 88:


	In the following picture, the first ROI is in the lake. You can see that the macro class ('''MC ID''') is named ''Water'' and the subclass ('''C ID''') ''Lake''. In this tutorial, only the macro classes will be significant, since it is a basic classification with only ~~four~~ different classes. If you want to have more specific classes you can use the subclasses. Save the ROI.		In the following picture, the first ROI is in the lake. You can see that the macro class ('''MC ID''') is named ''Water'' and the subclass ('''C ID''') ''Lake''. Make sure to change the '''MC ID''' name when creating ROI's for different land classes. In this tutorial, only the macro classes will be significant, since it is a basic classification with only five different classes. If you want to have more specific classes you can use the subclasses. Save the ROI.

← Older revision		Revision as of 17:09, 2 October 2021
Line 59:		Line 59:


	After defining the section under '''Clip coordinates''' there should ~~occur~~ ~~numbers~~. Click run and define an output folder. In the '''Layer Dock''', for each Band (1-9,11,12) a separate resized Raster Layer occurs. Once your clip layers have been created, you may remove the original band layers to free up some space in your layers legend.		After defining the section, under '''Clip coordinates''' there should be coordinates for the area you just defined. Click run and define an output folder. In the '''Layer Dock''', for each Band (1-9,11,12) a separate resized Raster Layer occurs. Once your clip layers have been created, you may remove the original band layers to free up some space in your layers legend.

	===Automatic Conversion to Surface Reflection===		===Automatic Conversion to Surface Reflection===

← Older revision		Revision as of 16:59, 2 October 2021
Line 155:		Line 155:


	Built-up area (brown line) and unhealthy vegetation (turquoise line) have very similar spectral signature plot and the algorithm uses these signatures for the calculation. However, you can reduce this error by setting more ROIs. Another possibility would be to include indices in the classification which are explained in the Tutorial mentioned above ([[Remote Sensing Analysis in QGIS]]). Unfortunately, you ~~can not~~ totally overcome the error.		Built-up area (brown line) and unhealthy vegetation (turquoise line) have very similar spectral signature plot and the algorithm uses these signatures for the calculation. However, you can reduce this error by setting more ROIs. Another possibility would be to include indices in the classification which are explained in the Tutorial mentioned above ([[Remote Sensing Analysis in QGIS]]). Unfortunately, you cannot totally overcome the error.

← Older revision		Revision as of 16:58, 2 October 2021
Line 167:		Line 167:

	===Calculate the Kappa Coefficient===		===Calculate the Kappa Coefficient===
	A quantitative method to assess the classification is to calculate the Kappa Coefficient. First, you have to create a new layer with ROIs and set again ROIs for the four classes to have a reference ground. You ~~can not~~ use the ROIs you used for the classification because you want to compare the classification with undependable training input. After you created various ROIs open the SCP and go to '''Postprocessing''', '''Accuracy'''. As your input layer choose your best classification result. The reference raster layer will be the new ROIs you just set:		A quantitative method to assess the classification is to calculate the Kappa Coefficient. First, you have to create a new layer with ROIs and set again ROIs for the four classes to have a reference ground. You cannot use the ROIs you used for the classification because you want to compare the classification with undependable training input. After you created various ROIs open the SCP and go to '''Postprocessing''', '''Accuracy'''. As your input layer choose your best classification result. The reference raster layer will be the new ROIs you just set: