You can use R programming language using R studio for classifying Landsat satellite image for which tutorials are available in youtube

If you want to work with numerical data, you can use WEKA which gives the RF along with some other methods.

But if your data are images, you can implement RF in MATLAB.

I suggest random forest classifier of scikit learn in Python language.

Depends on what you'r most familiar with.

Land use classification using GEE examples are available in abundance; using Google Colab (where you install scikit ML package) is also very user friendly and examples are also available - specifically on Github.

Good luck :)

I would look to Prof Qiusheng Wu's tutorials etc on Github if you're a beginner.

Dear Hossain Mohammad Arifeen,

You can use Orfeo Toolbox in QGIS.

You can use python.

And you can find some code about the random forest algorithm on github.

All software are good but depending on your specialization related on software. For me you should use Google Earth Engine.

Dear Hossain Mohammad

You can use various software. It depends on you to select one of them.

From my point of view, the "randomForest" package in R and Google Earth Engine (GEE) are good options for you. You can easily find related commands and scripts for both of them by a simple search on google.

Overall, I suggest you use GEE to develop classification models since you can visualize and monitor your reference points (test and train) at the same time.

Regards,

Ardalan

Google earth engine

try scikit-learn

Hi,

You can use the Google Earth Engine web_based platform.

surly GEE Platform

Hi Hossain,

You can use the R software and randomForest package or caret package. Hossain don't hesitate to contact me if you need other clarification.

here is the step-by-step r code:

https://valentinitnelav.github.io/satellite-image-classification-r/

also:

in SNAP software and Qgis, I think in new updates there are RF addins

I have tried both, the R and Software work well but the point is that your control on R code is higher than software and you can look very deep into your results but in Software you cannot manipulate RF too much out of the Software design.

Dear Arifeen,

This is a good question, you can do it easily with R software.

Just you need to install the "randomForest" package.

Hi..You can use the R software with randomForest or caret packages...