anggik

salam kenal anggi, geografi ui 2006

just share my experience once I ever did the same task for palm counting from quickbird imagery. I only use the red band (green also can) and did some supervised classification with proper tree sample of course. after that, simply convert the result to shapefile and did some conversion to centroid point. my result is about at 94% accuracy. but it only easy to run on palm with not very dense canopy. good luck

hi mettus do you have the tutorial from the spectro producer? I think you will get the proper method there for ground survey by using those product

I have try with ermapper by using standard classification (non supervised) but it only works on immature palm. for mature palm which have dense canopy it cannot work at all.