Hello,
We are mapping a winter wheat experiment in an attempt to evaluate several vegetation indexes. We have used a Sequoia camera and followed the documented procedure to generate a reflectance calibrated orthomosaic (
www.agisoft.com/pdf/PS_1.4_(IL)_Refelctance_Calibration.pdf). The orthomosaic was saved as a geotiff image without compression.
A similar procedure using the same photos and region of interest was done in pix4dmapper. The pixel values of the reflectance calibrated map generated by pix4dmapper are recognized as a floating point number (single) between 0 and 1 by geotiffread() when importing data to Matlab.
The issue we are facing is that the orthomosaic generated by Photoscan must be normalized after importing to Matlab using geotiffread(). However the normalization value is not documented. After importing to Matlab with geotiffread we are getting pixel values in 16 bits unsigned format.
If we divide each 16-bit pixel value by 65536 in order to obtain a reflectance value between 0 and 1, the resulting orthomosaic histogram is quite different (half value) from the one obtained from pix4dmapper.
What is the correct normalization factor in order to convert the 16-bit number associated with a pixel in the calibrated orthomosaic to a reflectance value between 0 and 1? Thanks.