I am currently working on the various indices that will allow me to clearly determine the built-up areas and the bare lands within a complex Moroccan region. As you know of course the index NDBI (Normalized Difference Built-up Index) for example is used to extract built-up features and indexes range from -1 to 1. It is calculated according to the following formula: NDBI = ( SWIR - NIR) / (SWIR + NIR)
That it will be worth the index built in the case of using a set of index like NDVI, MNDWI and NDBI.
I want to have maximum interaction in this topic.
Thanks in advance.
Hi,
Bare surface and built-up area have similar spectra properties. However, adequate groundtruthing could help you train your model for classification eg in case of machine learning.
Normally, the built up area should have higher magnitudes of negative values when you use ndbi.
Please,forget the ndvi in this case because of its weakness in separating moisture from soil. Msavi2 performs better in this case. Mndwi on the other hand is good for water extraction. But awei is better than the Mndwi in most cases.
Maybe this will help : https://www.researchgate.net/publication/318027710_Investigating_Surface_Urban_Heat_Island_Characteristics_over_Abuja_Nigeria_relationship_between_land_surface_temperature_and_multiple_vegetation_indices
Dear why you go through the false color to do the extraction by classification
Object based image classification, is one of the best method for delineating built-up and urban areas. Good luck.
- Badges
- Science topic
- Similar topics
- Geoscience
- Remote Sensing