Hello all
I want to estimate forest tree biomass-carbon using Landsat images. Plot-wise biomass-carbon has been calculated and standardized (to t/ha) for 124 field plots. Now, I need to establish the relationship between field carbon with spectral response of Landsat data. You ignore any statistical consideration as that is not my main concern here. Most studies used regression analysis of field data with image bands and vegetation indices. However, I need to use k-NN algorithm as it is non-parametric in nature and highly recommended for use in any deforestation-related emission estimation project for REDD+. Most people used programming languages to run kNN. I don't have sufficient knowledge of R/Matlab/Python coding and not having enough time to invest on excelling on that at the moment. So wanted to use kNN in a commercial RS GIS software like IDRISI/TerrSet, ENVI etc. Do you think it is possible? If yes, could you be kind to describe the procedure in a step-by-step manner? Please note, I don't need to get forwarded any theoretical document(s) on kNN as I have studied the available internet stuffs quite fairly. I need the the details of the (software) working procedures. I would appreciate your urgent responses please. Thanks!
Hi Mohammd,
I am sorry, I am not really understanding your question here. Can you elaborate a little on the format of your field data you want to link?
If you are talking about images- you can use ENVI- Layer stacking, and later apply KNN classification.
Otherwise you can use arcGIS to create a map of the field data you have and then later on linking the Landsat image as a raster to the map.
- Badges
- Science topic
- Similar topics
- Geoscience
- Geoinformatics
- Geographic Information System