What kind of image classification is suitable for any land use or land cover types, spatially in different climates?
Dear Saleh,
If you are using multi-temporal images of the same area for land cover mapping, you should also consider conducting atmospheric correction.
It is always better to work with radiance or spectral reflectance instead of DNs.
DN is the digital number that is generated in the imaging system of the meteorological or earth observation satellite. It is just number. DN values vary from pixel to pixel of the image depending on what features the imager is scanning. DN values are processed to find different parameters regarding the Earth's atmosphere or surface. For example DN value for bare soil may be lower than cloud (depends on type of sensor; visible infrared etc).
Dear Ali thank you so much for your attention on my question. actually i compared 6 image classification methods to produce land use map for tow climates. results showed that for Dry climates ML is better than other methods but in Humid climates SVM is better than other. you know, i don't understand what is the effect of climates on image classification algorithms.
i will send you the manuscript if you like
thank you so much.
Dear Saleh,
If you are using multi-temporal images of the same area for land cover mapping, you should also consider conducting atmospheric correction.
It is always better to work with radiance or spectral reflectance instead of DNs.
- Badges
- Science topic
- Similar topics
- Geoscience
- Atmospheric Sciences
- Climate Science
- Climate Change