Nasir Hameed Thank you for your valuable response.

It depends upon your interpretation which is a key to accuracy and can be improved using indices and various online extraction tools like Settlements from OSM.

Aazim Yousuf Thank you for your response.

You can utilize a series of software to this end but personally speaking ENVI works wonders in this regard

Behnam Khorrami Thank you for your response.

Generally speaking, given the diverse range of climate systems and terrain conditions in different areas, there is no conclusive agreement on introducing a specific approach that would provide the most accurate results in all circumstances. However, RF and SVM usually show acceptable performances. I would recommend you to read following papers to get more information (codes are also available):

Article A review of land use/land cover change mapping in the China-...

Article RUESVMs: An Ensemble Method to Handle the Class Imbalance Pr...

Article A Hybrid Data Balancing Method for Classification of Imbalan...

Amin Naboureh Thank you for your response. I will definitely gone through those papers.

ERDAS Imagine and ENVI (SVM) methods is better

Ismail Mondal Thank you for your response.

Most GIS and RS software have the standard classification methods (ArcGIS, QGIS, ENVI, Eedas etc).

Dear Pritiranjan

I would check out the Queensland, Australia, SLATS website. https://www.qld.gov.au/environment/land/management/mapping/statewide-monitoring/slats . The people running this are experts and no doubt will be happy to answer questions directly.

Nabanita Mukherjee Thank you for your valuable response.

Dear Pritiranjan Das

Most of the Geographic Information Systems programs have good classifiers, it's a matter of testing and seeing if you get the result you want (ArcGIS Pro, QGIS).

On the other hand, you have programs that were created for digital image processing (PCI Geomatics, ENVI, TerraAmazon), in which you find more options. These programs also have the basic GIS tools.

Finally, the trend is for these processes to be carried out in the cloud to streamline processes, here a very good tool is Google Engine.

Note: My program references are based on what I have used, surely there are other programs (possibly with better results)

Andrés Velástegui Montoya Thank you for your valuable response.

Hi Pritiranjan Das,

You can use Google Earth Engine for LULC Classification using machine learning algorithm. You can easily find the code for that in GEE itself. You just have to give the training data sets according to the Landsat imageries. Most importantly, try different band combinations for identifying the classes keeping the base map as reference. It will take less time for such classification. You can even check the training accuracy and validation accuracy with the preparation of the LULC map using GEE.

Sushree Sangita Dash Thank you for your valuable response.

Hi Pritiranjan Das,

Im using ERDAS for classification and using google earth imagery for comparison

Dear Pritiranjan Das ERDAS Imagine is a perfect image classification software. I agree with others that all GIS software (Open or licensed) have standard image classification algorithms but ERDAS Imagine is exceptionally made for image classification. So, I suggest you do image classification using ERDAS Imagine. There are options available, you can do either supervised or unsupervised classification based on the applicability of your research. I prefer supervised image classification using the maximum likelihood classifier for my studies.

Gibji Nimasow Thank you sir for your valuable response, i must follow your preference.

Regards.

I suggest you to use Google Earth Engine.

Please refer the link below for LULC classification using Random Forest

https://geohackweek.github.io/GoogleEarthEngine/05-classify-imagery/

You can also use different ML algorithm for the classification in GEE easily.