automatch

A python tool for automatic image georeferencing.

options

-i input image -r reference (georeferenced) image -o output image --gcps if present, outputs the gcps as a shapefile next to your output image (optional) --feature choose the feature detector. At the moment, the following are allowed: <sift|surf|orb|akaze|brisk> --flann if present, use the FlannBasedMatcher. Otherwise, use the BFMatcher (Brute Force) -tileX x dimension for the tiles (optional) -tileY Y dimension for the tiles (optional) -offsetX x dimension for the offset between tiles (optional) -offsetY Y dimension for the offset between tiles (optional)

example

python3 automatch.py -i your_awesome_image.jpg -r an_already_georeferenced_map.tif -o your_new_awesome_georeferenced_image.tif --gcps -tileX 512 -tileY 512 -offsetX 512 -offsetY 512

Precomputation of keypoints and descriptors

Alternatively, you can save the keypoints and desctiptors as a custom file (sort of project file). All the arguments used for the point detection will be saved.

Usage

Usage: save_keypoints.py [OPTIONS] INPUTFILE OUTPUTFILE

  Compute the keypoints and desctiptors of an image and save them in a store
  file. All the arguments used for the point detection will be saved.

Options:
  -f, --feature TEXT             <sift|surf|orb|akaze|brisk>[-flann]
  -n, --n_features INTEGER       Number of features
  --tile <INTEGER INTEGER>...    Tile dimensions (width, height).
  --offset <INTEGER INTEGER>...  Offset as a pair (offsetX, offsetY)
  --help                         Show this message and exit.

Example

python save_keypoints.py  some_image.jpg  store.pickle --tile 512 512 --offset 512 512 --feature orb

Visualisation of keypoints and descriptors

Now that they have been saved, you can visualise your keypoints store in the custom file (sort of project file).

Usage

Usage: load_keypoints.py [OPTIONS] KEYPOINTS

  Parse the load keypoints arguments

Options:
  -n, --n-keypoints INTEGER  Number of keypoints to load
  --help                     Show this message and exit.

Example

python3 load_keypoints.py store.pickle

Run automatch with the stored keypoints and descriptors

You can also run automatch with the stored keypoints. That reduces the number of required arguments:

options

-ki input project file -kr reference project file -o output image --gcps if present, outputs the gcps as a shapefile next to your output image (optional) --flann if present, use the FlannBasedMatcher. Otherwise, use the BFMatcher (Brute Force)

example

python3 automatch.py -ki your_awesome_project_file -kr an_already_georeferenced_map_project_file -o your_new_awesome_georeferenced_image.tif --gcps --flann