I face a problem in LULC classification, like as an industrial area showing as a water body.
Please help me.
Thank you.
You can use combination of indices to classify phenomena more accurately. I suggest using Modified Normalized Difference Water Index (MNDWI) and Normalized Difference Vegetation Index (NDVI) for more accurate classification of water. In the infrared band, water has very little reflectance compared to the red band. As a result, in NDVI index, negative values will indicate water bodies. NDVI is as follows:
NDVI = ( Near-infrared - Red)/(Near-infrared + Red)
For Landsat 8, NDVI is as follows:
NDVI = ( Band 5 - Band 4)/( Band 5 + Band 4)
For Landsat 5 and 7, NDVI is as follows:
NDVI = ( Band 4 - Band 3)/( Band 4 + Band 3)
In more densely populated urban areas the value of normal NDVI is positive, although closer to zero. Water bodies which have a rather low reflectance result in slightly negative NDVI values.
You can apply pan-sharpening process on your satellite image (on Envi Software) with using panchromatic band. Then you can increase your resolution from 30m to 15 m. So you can get higher accuracy results with good classified LU and LC image.
Yurii Moskalenko, Nasir Hameed , Afshin Amiri and Nursaç Serda Kaya I am thankful to each of you for your valuable comment.
Dear It has the QGIS program. Through the bands it is possible, you only have to take care of the leftover clouds
Try to get additional training data and to identify threshold values for your land cover classes. You can also try Otsu method (in Envi) for a better separation of land cover classes.
Vasilică-Dănuț Horodnic Thank you for your suggestion; can you suggest me any research paper about this method?
Pritiranjan Das, you could try these:
1) https://www.youtube.com/watch?v=T5hV_5LGos4
2) A new spectral index for the extraction of built-up land features from Landsat 8 satellite imagery
- Badges
- Science topic
- Similar topics
- Geoscience
- Remote Sensing