3

I need advice on how to complete the following task:

I have aeronautical charts covering parts of the US (example: http://aeronav.faa.gov/content/aeronav/sectional_files/Albuquerque_89.zip - Each zip contains ~2 files covering both the northern and the southern part of the area). Unfortunately for me, these maps have collars that I want to remove so that I could create a new geotiff file with only the real content of the original geotiff. I have tried Global Mapper's "Crop collar" command, but that didnt gave me the results I wanted to have. In my understanding I should use a tool like GDAL warp to complete this task in bulk for all of the images (~106 images in total), but the situation is that I am completely inexperienced using that tool. I have already tried running gdalwarp -srcnodata 0 -dstalpha input.tif output.tif but that didn't gave me good results as well, therefore I would be glad to receive some help!

Thanks in advance

Improve this question
3
  • If you could produce a polygon for your collars, gdal's cutline may be of interest to you. Also try running gdal's nearblack command before executing the gdalwarp to help eliminate the collars. – Peter Smith Jul 12 '12 at 15:07
  • Yeah, but thats the issue, as I really don't want to spend time with going thru all the maps and define those polygons by hand. Do you think there is any other workaround to solve this problem? I was thinking about identifying NODATA fields and get rid of them automatically (via some kind of cropping). Maybe the collar pixels are NODATA, but I am not sure as I don't have that deep knowledge as I want. ;) – EZMapdesign Jul 18 '12 at 13:31
  • Did you ever find a solution to this question? I have similar charts and would like to know how, if you were able :) – Nick Badal Oct 23 '13 at 21:05
1

There are a few ways to do this. One is through arcCatalog 10.1 and the new mosiac datset option. Under the Product Properties drop down there are options to select each band and the min/max values for each band.

Identify what the highest and lowest pixel values are for the data that you do not want. List them for each band.

The second option is to use FME. Using a fairly complex workbench I was able to create a polygon shapefile based off of each pixel value using the equation 

if(A[0]<$(MAXPIXVALUE)&&A[1]<$(MAXPIXVALUE)&&A[2]<$(MAXPIXVALUE),0,A[0]);if(A[0]<$(MAXPIXVALUE)&&A[1]<$(MAXPIXVALUE)&&A[2]<$(MAXPIXVALUE),0,A[1]);if(A[0]<$(MAXPIXVALUE)&&A[1]<$(MAXPIXVALUE)&&A[2]<$(MAXPIXVALUE),0,A[2])

Where MAXPIXVALUE is a user parameter set to the max pixel value to be considered black. This equation searches each band in each pixel and creates the polygon, which can then be used to clip away the imagery that you want.

Improve this answer

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.