Changes between Initial Version and Version 1 of Ticket #2048, comment 27


Ignore:
Timestamp:
Mar 5, 2019, 7:49:53 AM (5 years ago)
Author:
cmbarton

Legend:

Unmodified
Added
Removed
Modified
  • Ticket #2048, comment 27

    initial v1  
    331. #774 is a discussion about support for >8bit values in i.rgb.his and i.his.rgb. I cannot tell from this discussion or from the module manual if this happened or not. I'm guessing it did not. Can anyone verify?
    44
    5 2. The main place where i.pansharpen is explicitly 8bit is in the histogram matching method. The main goal of pan sharpening, AFAICT, is to produce an RGB image that is visually pleasing at spatial resolutions equivalent to a high res pan band. The original 'data' are lost regardless of the bit depth. Given this, the most straightforward solution to #1 and #2 is to transform the original 4 bands (RGB and pan) to 8bit before doing any processing. Any reason not to pursue that strategy?
     52. The main place where i.pansharpen is explicitly 8bit is in the histogram matching method. The main goal of pan sharpening, AFAICT, is to produce an RGB image that is visually pleasing at spatial resolutions equivalent to a high res pan band. The original 'data' are lost regardless of the bit depth. Given this, the most straightforward solution to both of these issues is to transform the original 4 bands (RGB and pan) to 8bit before doing any processing. Any reason not to pursue that strategy?