Dear experts, I badly need a help to solve the problem of Landsat image classification of a coastal region of Bangladesh. The problem is regarding supervised classification of a coastal district. The land cover spectral reflectance value of pixels is quite complex. The Pixel values of Build up area and Barren land are very close. When selecting the Training sample pixels for Urban area and Barren lands that are often in conflict in the output. There are many unwanted areas in the classified image. Like many unwanted buildup areas or many Barren lands occupy Buildup areas. I tried more and less training inputs for urban area several times. but the results are not appropriate. Can you please suggest any ways to solve the problem??
Hi Md Sarowar Hossain ,
What about the infrared bands? Is it possible to distinguish barren lands and buildup areas by visualising the spectral signatures of near infrared and SWIR? Is there still a confusion? Perhaps, you can also make a mask of both classes and apply a new classification on it or a Principal Component analysis to dissociate them.
Best regards,
Rémi DM
Dear Rémi De Matos-Machado
I tried different band composition. But those didn't give a large appropriateness in the visualization of composite image. I said the land cover classes are quite complex. But as your suggestion I can try different algorithm. I will check principle component analysis and will let you know about the result.
Md Sarowar Hossain,
So if understand, you made your samples using the composite image and then made a classification based on the values of the composite image (not the bands directly)? If yes, you should use the bands directly (and especially infrared bands).
Best regards,
Rémi DM
- Badges
- Science topic
- Similar topics
- Computer Science
- Images