@@coolsamuel4354 I am using SVM Classification method. it's only provide you the RGB count values. I think you should apply the method then see the results.
sir one doubt i have using sentinel 2 5 band composite imagery for my project, can I train rf using training samples generated from that sentinel 2 5 band imagery for current year and save the .ecd file and use it for same area but different date imagery with same 5 band composite imagery for image classification whether it will work please help me sir
You can use it but Result will not be accurate . Because you have made training samples using different imagery. If you think your training samples are exactly same as the previous image then you can try it. Look at the final result after using it.
Thanks for sharing this informative SVM Tutorial. I had one question, How can we do the accuracy assessment of this lulc, like User and Producer Accuracy, Kappa Coefficient and Root Mean Square Error (RMSE)?
Hello, thank you for such a great video. It helped me a lot. I have a question. I have created a training shapefile based on a Landsat 9 image. It works well at classifying the Landsat 9 raster. However, I want to also classify the same area but using a Landsat 7 image from 20 years ago. Is there a way I can use the same training data for the Landsat 7 raster? When I try to, I get the following error: "Band numbers in input rasters differ from band numbers in classifier definition." I'm wondering if you know a way around this. Thanks!!
I am getting your point. Simply just run the tool "Train SVM" same as video. Landsat 7 will be your Raster Image & training sample will be your Current Landsat 9 training samples. Run it. It will work.
@@geoinformaticsintelligence thank you for your concern sir.... I'll be looking forward for the video. I was able to do the accuracy assessment using Create Accuracy Points and Confusion Matrice tools in the arctool box... 😇😇😇
Goodwrk, I have been stuck while doing Supervised classification. I have made trainning set using simple digitization in Qgis and I am trying add shapefile to arc to make Signature file, Can you help me on this
Sorry forget my last comment, I will guide you. First you should change the field attributes names of that signature file. Add Two Fields - classname (text), classvalue(Long integar). you can take example from my this video. 1. Go to ArcGIS Search (Train Maximum likelihood classifier). Add your signature file & given the path where you can find .ECD file easily. 2. Now Search (Classify Raster). add Satellite Image & .ECD file of signature. If you still finds some difficulties then you can whatsup me on my number 7529918160.
Informative video ❤
Glad you think so!
Excelente video…!!!!
Thank you 😊
Nice
Thanks
Excellent tutorial again, but I also noticed that the output doesn't have pixel values in its attributes
Yes, you are right, but why do you need for a pixel values?
Classification have only provide you the count values of pixels for each LULC class.
@@geoinformaticsintelligence there's no pixel value for each corresponding class
@@geoinformaticsintelligence sorry I mean pixel counts
@@coolsamuel4354 I am using SVM Classification method. it's only provide you the RGB count values. I think you should apply the method then see the results.
sir one doubt i have using sentinel 2 5 band composite imagery for my project, can I train rf using training samples generated from that sentinel 2 5 band imagery for current year and save the .ecd file and use it for same area but different date imagery with same 5 band composite imagery for image classification whether it will work please help me sir
You can use it but Result will not be accurate . Because you have made training samples using different imagery.
If you think your training samples are exactly same as the previous image then you can try it.
Look at the final result after using it.
@@geoinformaticsintelligence Thanks sir
Thank you so much! But, could you please tell me which satellite type did you use for this classification?
I have used Sentinel 2 satellite image.
Thanks for sharing this informative SVM Tutorial.
I had one question, How can we do the accuracy assessment of this lulc, like User and Producer Accuracy, Kappa Coefficient and Root Mean Square Error (RMSE)?
There is tool in arcgis to calculate your accuracy. you can search on google.
Hello, thank you for such a great video. It helped me a lot.
I have a question. I have created a training shapefile based on a Landsat 9 image. It works well at classifying the Landsat 9 raster.
However, I want to also classify the same area but using a Landsat 7 image from 20 years ago. Is there a way I can use the same training data for the Landsat 7 raster? When I try to, I get the following error: "Band numbers in input rasters differ from band numbers in classifier definition." I'm wondering if you know a way around this.
Thanks!!
I am getting your point. Simply just run the tool "Train SVM" same as video.
Landsat 7 will be your Raster Image & training sample will be your Current Landsat 9 training samples.
Run it. It will work.
Pla also tell hiw to do accuracy assessment of a LULC that have been made by svm method..
I will make a video on it. wait for the update.
@@geoinformaticsintelligence thank you for your concern sir.... I'll be looking forward for the video.
I was able to do the accuracy assessment using Create Accuracy Points and Confusion Matrice tools in the arctool box... 😇😇😇
@@anubhavmaurya8959 Sure
How about accuracy assessment
Do you know? How to do accuracy assessment?
Excellent tutor. Av been stuck.I can find the .ecd file in file manager but arcgis can't locate the same . kindly help on this
You can save .ecd file in a folder where you can access & find easily into (D drive).
I am unable to locate .ecd file after keeping it in proper place please help
It's ok. I understand.
You can save ecd file on desktop then you will easily find your .ecd file.
It will work. Do it
Goodwrk, I have been stuck while doing Supervised classification. I have made trainning set using simple digitization in Qgis and I am trying add shapefile to arc to make Signature file, Can you help me on this
You can not create signature file from external format. You should do classification in Arcgis.
Sorry forget my last comment, I will guide you. First you should change the field attributes names of that signature file. Add Two Fields - classname (text), classvalue(Long integar). you can take example from my this video.
1. Go to ArcGIS Search (Train Maximum likelihood classifier). Add your signature file & given the path where you can find .ECD file easily.
2. Now Search (Classify Raster). add Satellite Image & .ECD file of signature.
If you still finds some difficulties then you can whatsup me on my number 7529918160.
Sir what is the main difference between Supervised classification and SVM
SVM, You can be used for both classification supervised or Unsupervised.
Density kernel…
Sorry, I am not getting your point.