[GIS] Seeking Software for Stitching Aerial Photographs

aerial photographysoftware-recommendations

I am looking for software that is capable of stitching together large number of aerial photographs. The photos are taken with a high quality DSLR from a small helicopter, they cover about 100 meters, they have a small amount of overlap, they also have location data stored in the exif header, are about 7 mb in size, and there are usually around 30000 of them. I was just wondering if anyone had any suggestions.

So far I have tried:

  • Autopano Giga 3 – no good results yet just some weird stitches

  • Palentier (yes I know it doesn't really stitch the photos) – Hard to use as I am lacking the direction info, and height above ground info in my exif headers. The height above ground can be calculated by subtracting the ground elevation from the elevation stored in the photo exif but all photos must have the same subtraction value and the terrain height varies greatly.

  • Agisoft Photoscan – Doesn't like the large amount of images, but I am currently trying it with smaller batches.

Best Answer

This question has been converted to Community Wiki and wiki locked because it is an example of a question that seeks a list of answers and appears to be popular enough to protect it from closure. It should be treated as a special case and should not be viewed as the type of question that is encouraged on this, or any Stack Exchange site, but if you wish to contribute more content to it then feel free to do so by editing this answer.

There is Open Drone Map, that calls itself "Open Source Toolkit for processing Civilian Drone Imagery".

You can run it in docker container and you will get your own aerial photography service in the cloud.

Or you could try DroneDeploy cloud service, you can upload your photos to their cloud and get up to 5 maps for free per month.

Besides just getting orthomosaic photo, you can also get elevation maps and 3D maps which will blow your mind.

ImageMagick

supports: Large image support: read, process, or write mega-, giga-, or tera-pixel image sizes.

command --line will the most efficient way to do this:

(via -layers method)

merge

As 'flatten' method but merging all the given image layers to create a new layer image just large enough to hold all the image without clipping or extra space. The new images virtual offset will preserve the position of the new layer, even if this offset is negative. The virtual canvas size of the first image is preserved.

Caution is advised when handling image layers with negative offsets as few image file formats handle them correctly. Following this operation method with +repage will remove the layer offset, and create an image in which all the overlaid image positions relative to each other is preserved, though not necessarily exactly where you specified them.
See also 'trim-bounds' below which is closely related but without doing the'flatten' to merge the images together.

mosaic

As 'flatten' method but expanding the initial canvas size of the first image in a positive direction only so as to hold all the image layers. However as a virtual canvas is 'locked' to the origin, by its own definition, image layers with a negative offsets will still become clipped by the top and left edges. See 'merge' or 'trim-bounds' if this could be a problem. This method is commonly used to layout individual image using various offset but without knowing the final canvas size. The resulting image will, like 'flatten' not have any virtual offset, so can be saved to any image file format.

http://www.imagemagick.org/script/command-line-options.php#layers

Most OS Platforms supported

ImageMagick 6.8.6-10 available from http://www.imagemagick.org/download

You might want to look at Microsoft ICE. Its free, but its not a geospatial solution. I have used it with good success. It will make a good mosaic, and then you can reference the mosaic.

The best (commercial) product I have encountered for mosaicing images is Erdas Imagine's Mosaic Pro (MP). There are several features of MP that I particularly like:

  1. MP utilizes all of the cores of your machine.
  2. Handles > 2.5 TerraPixels and thousands of images
  3. There are a variety of seamline and overlap rule options available
  4. There is a good variety of output formats
  5. You can create new subtiles based on a shapefile
  6. You get access to the rest of Erdas's functionality
Related Question