Hello Kuria
The zero values that you mentioned is independent of RapidEye data since you can get those values for Landsat Imagery as well. If you meant such as classification, you can manipulate/ delete those values or group those values as "unclassified" then remove it from your results. Hope I could understand it correctly.
Also please follow the link below and read Onur Satir's reply
https://www.researchgate.net/post/How_can_I_change_0_value_pixels_to_NoValue_in_Erdas_Imagine/1
Typically images do not have a DN value of zero; reflectance wise, there is nothing on the surface of the Earth the is absolutely black / zero percent reflectance. If after haze corrections a 'few' pixels could be set to zero but I have suggested in on of my radiometric calibration and atmospheric correction paper (1989) that a more realistic min reflectance is 1%. Pixels with zero DN value are typically background / border areas and zero SHOULD NOT BE USED because it can skew statistical analyses results dramatically.
Pat
One more useful scenario to consider:
Consider the geometrically corrected/warped image read as raw data. The image will be rotated/warped and will be embedded inside a large rectangle filled with zeros. But those zeros have been artificially introduced by your algorithm. Use this option to ignore the zeros and get the statistics after your algorithm has processed the input, to compare the degree of distortion your algorithm has introduced.
- Badges
- Science topic
- Similar topics
- Biological Science
- Bioecology
- Systems Ecology
- Landscape Ecology
- Connectivity