Supervised classification in Google Earth Engine

Guided tutorial providing an introduction to supervised classification of Landsat 8 imagery in Google Earth Engine.
www.geospatiale...
Assoc. Prof. Shaun R Levick

[@bhanukhanal7662]

so only white buildings will be considered as urban? what about buildings with other colors? please reply, thanks

[@lostadv]

Very informative -- thank you Dr. Levick.

[@gears-geospatialecology]

Thanks Dr. Miller, good to hear from you.

[@viveksingh-vr1fi]

Thanks, Sir
It is really helpful for us.
Can you please tell me how to implement Neural Network in 'Sentinel 2 MSI' data. using the google earth engine?

[@andysaputra778]

good explanation sir, and now can you export the processed data into a processed file so that no more reprocessing is needed?

[@neeturathi6483]

Thank you Dr. Levick. really helpful. Can you please tell about how we will apply the different classifier and calculate the area of the classified image.

[@gears-geospatialecology]

Hi neetu rathi - I'll be posting some fresh material in the next few weeks.
Regards,
Shaun

[@thrasivoulosstylianou3567]

Hello, I want to make a classification in Sentinel 2 Image. I want to use all bands. Due to the different spatial resolution, I can not do the classification in this way. I had read, that I can do the resampling process (nearest neighbor ) by contacting the reprojection process. However I don't find how to do it. Can you please help me ?
Thanks in advance !

[@eldelbharadri2282]

Hi there!
When using this code, I am getting the following error: (Line 10: urban.merge is not a function).
Could it be because .merge is not a function anymore and nowadays another function has been developed? How could i solve it?

[@thenomadicvoice16]

I want to mark ROI as a rectangle and then I want to take the feature collections as polygon covering the boundary of the ROI area. So can you help me with this? Thanks!

[@arisvictorjr.luarca1336]

hello! it says smilecart cannot be read properly. is there a way to remediate this?

[@nandanabn6477]

Sir , please post video for to use polygons instead of points for training data in supervised classification? I am hardly in need of the above same. Waiting for your video

[@giannisps1111]

Hello, has GEE any issue on usinng geotiff images as for the training data? It always exceedes the memory the server allows me to use.

[@bhanukhanal7662]

this was pixel-based, How to perform object-based classification with Landsat image on google earth engine? please make a video on it

[@chathuradesilva400]

Thank you very much for the video. I want to know how we can use the classification for time series analysis. For example how 'urban' had changed over the period. Can some one help me on this!

[@apurbasarkar8064]

You can use autamata

[@nisha8504]

Where from I can get this code?

[@shreearmygirl9878]

sir pl can u suggest what is this error is 2020: Layer error: Property 'landcover' of feature '1_1_1_1_0_0' is missing.

[@himpria]

Thanks for your informative video,
I have one query,

[@efosaadagbasa6895]

Hi Levick, please do you know how to use ANN classifier with GEE ?

[@erickodire8006]

Hello sir, I got an error
"Later error: classifier.cart: this classifier has been removed"
Kindly advise, thanks

[@krishankumarsanodiya7682]

Use ‘classifier.smilecart’

[@bhanukhanal7662]

add smile

[@endalkachewgoshedagnew9123]

you did great sr, thank you

[@amirarasyidah1040]

can i do maximum likelihood in GEE?

[@noufiama209]

Sir, Can you please explain section 3, spectral unmixing?

[@coolsamuel4354]

Hello, how do I export bcoz I kept encountering error messages

[@nandanabn6477]

please provide the video for unsupervised classification

[@arpanachaudhary2763]

when i am doing the last step it is showing this type of error "classification: Layer error: Property 'landcover' of feature '1_2_0_0' is missing" please tell me what to do now

[@СарварбекЭльтазаров]

I have similar error: Layer 1: Layer error: Property 'landcover' of feature '0_0' is missing. Could you solve it?

[@efosaadagbasa6895]

Very nice. Good Job

[@kc3928]

Hi Prof. Levick, would it be better to use polygons instead of points for training data? If you could reply today or tomorrow it would be greatly appreciated as I need this information as soon as possible thanks!

[@gears-geospatialecology]

Hi K Clair
Yes it would be better to use polygons - you would capture more variability within a class and have better sample sizes for training and validation. The process is slightly more complicated but is explained nicely here:
developers.google.com/earth-engine/classification
Regards,
Shaun

[@dilkumarbantawarai736]

In the context of Nepal, the Landsat imagery have a lot of linear gaps that affecting image classification. How to fill the such linear gaps using GEE plateform?

[@bhanukhanal7662]

use Landsat 8

[@tungminh6229]

Very nice, Thank you !

[@名字-m2m]

Very nice, Thank you.

[@dogimate]

Hi Levick. How do I export this classification to .tiff image

[@efosaadagbasa6895]

// Export the image to your google drive, specifying scale and region.
Export.image.toDrive({
image: classified,
description: 'classification',
scale: 30,
region: classified.clip(roi)
});

[@AkbarAli-vz9mn]

Hi@@efosaadagbasa6895 i want to export the classified image to google drive but the resultant exported image is in many tiles and disrupted with many grid lines but its not required result, how can i resolve my problem.
please reply asap

[@coolsamuel4354]

@@efosaadagbasa6895 Pls the exported file came in about 2kb and had one one layer and doesn't work

[@geospatial_classroom]

Thank you

[@joseluizalves7885]

Ola seria possivel mantar na linguagem brasileira ficava melhor o nosso entendimento

[@geospatial_classroom]

@@joseluizalves7885 srrry dnt understand

[@tsegayaregawi9529]

it is very interesting

[@danielvilardasilva9992]

Very nice, Thank you!