06-24-2019 08:39 PM
I am having an issue with using the HPF Resolution Merge tool. I am working with landsat 8 OLI tif files. I input Band 8 (15m res) for my high resolution file and then a stack file of bands 2-7 (30m res). The files that I am working with have just undergone a atmospheric correction from using QGIS. I am using QGIS Semi-Automatic Classification Plugin over the ERDAS ATCOR2 becuase of the quality (it's just a bit nicer). I am not sure if this could be a reason for the bazar output. I am using ERDAS 2014. I've tried running "Modified IHS Resolution Merge, HCS Resolution Merge, and Resolution Merge", but I get the same output each time (white canvas).
Image one shows bands 2-7 with a multiple-spectural output of 6, 5, 2
Image two shows band 8
Image three shows the parameters I used in ERDAS HPF Resolution Merge (left them as default)
Image four show the output I am recieving.
Solved! Go to Solution.
06-24-2019 05:41 AM
...and what does Inquire Cursor say the DN values are in the image area? It may be that you need to recalculate statistics ignoring the "background" area value.
07-01-2019 06:10 PM
Sorry I have been offline for the past week. Yes you are right, there is somthing going on with my DN values that were calculated through the QGIS atmospheric correction. Below are DN values for:
Are you suggesting recalculating stats on my QGIS atmospheric correction or the QGIS atmospheric correction with ERDAS pan_sharpening file?
07-02-2019 10:51 PM
Sorry Jonnie, I didn't see your post. Below are four pictures showing four of the bands (2, 3, 4, 5). It looks like their mean value never really change.
07-17-2019 02:28 AM
it’s possible that the HPF Merge algorithm expects an unsigned integer data range as input. To test this please try the following experiment. Multiply your atmospherically corrected data by, say, 10,000 and write a new output file out as a u16 data type. You can do this easily using Spatial Modeler. Then try merging that data and see if the results look better.
07-19-2019 07:56 PM
I don't know how, but I did end up getting the Pan-sharpening tool to work. I am no longer getting a pure white output. I no longer have the files that I was using that were outputting as the white canvas, so I cannot tell you if it was in a float data type. Though, I am having trouble with your Mosaic function. I will post a new question for this.