I have about 30 or so composite rasters in IMG format that I need to 1.) Preserve the date that the raster was created and 2.) Combine all of them. My workflow is:
Step 1: raster calculator, multiply floating raster by 10 so that decimal values are not lost in integer conversionStep 2: convert floating raster to integerStep 3: create attribute tables for the rastersStep 4: Add a fieldStep 5: Calculate field to add the date which will be taken from the name (the datetimes are in the names)Step 6 create raster mosaic - make sure it is still an integer.
This seems easy enough, but I'm not the best informed when it come to rasters. After step 2 (or step 1 if I multiply it by an integer) I get the raster with an attribute table, but it has lost its RGB coloring scheme. I thought that perhaps I could assign RGB based on pixel values as a step 7 once everything is done (somehow) but the pixel values do not quite match the RGBN individual bands. The range of the values of the new integer dataset is 3760-16180 (meaning the range was 376-1618 before Calculate Raster), which are the same values as the first band only (Red I think) and the composite stretched value range. Why is this? Why does the composite band use only these values as well when there are values lower and higher than this in the other bands? How can I classify the pixels as RGB if the full range for each is not there? Also, if you have alternatives or suggestions to my workflow, please let me know.
أكثر...
Step 1: raster calculator, multiply floating raster by 10 so that decimal values are not lost in integer conversionStep 2: convert floating raster to integerStep 3: create attribute tables for the rastersStep 4: Add a fieldStep 5: Calculate field to add the date which will be taken from the name (the datetimes are in the names)Step 6 create raster mosaic - make sure it is still an integer.
This seems easy enough, but I'm not the best informed when it come to rasters. After step 2 (or step 1 if I multiply it by an integer) I get the raster with an attribute table, but it has lost its RGB coloring scheme. I thought that perhaps I could assign RGB based on pixel values as a step 7 once everything is done (somehow) but the pixel values do not quite match the RGBN individual bands. The range of the values of the new integer dataset is 3760-16180 (meaning the range was 376-1618 before Calculate Raster), which are the same values as the first band only (Red I think) and the composite stretched value range. Why is this? Why does the composite band use only these values as well when there are values lower and higher than this in the other bands? How can I classify the pixels as RGB if the full range for each is not there? Also, if you have alternatives or suggestions to my workflow, please let me know.
أكثر...