Dear all, 

I am classifying Land use land cover map based on NDVI value (category for classification : Agriculture area, Build up area, Forest Area, Water body, Barren land and grass land). I am not running any supervised or unsupervised classification or other classification techniques. cause some classes not classified well and its gets mix with another class and recoding of those classes will not probably give accurate results. So i am thinking to do indices based (NDVI, NDWI, NDBI) classification  to generate the LULC map.

I am using multi temporal LANDSAT images (Ravi, Kharif and summer season).   

My confusion is only indices based feature extraction for classification is scientifically correct or not ? or i must have to run any classification techniques. 

Please share your opinion and experience with me.

Thank you

Regards

Shouvik Jha  

Similar questions and discussions