16.36. ipmatch

The ipmatch program reads interest points (IPs) from .vwip files produced with ipfind (Section 16.35) and finds matches among them, writing out .match files containing the results. ipmatch can also produce debug images with plotted match points.

If more than two image/vwip sets are passed in, each possible combination of images will be matched.

The produced binary match files can be visualized in stereo_gui (Section 16.67.9) or converted to plain text for inspection with parse_match_file.py (Section 16.51).

Note that this tool does not implement many of the IP matching steps that are used in parallel_stereo and bundle_adjust, since it does not use any camera information.

Example:

ipmatch image1.tif image2.tif image1.vwip image2.vwip

The order of images and .vwip files is flexible. The first image will use the first .vwip file encountered, etc. The .vwip files can be omitted altogether, and then can be deduced from image names. So the program can be also called as:

ipmatch image1.tif image2.tif

Usage:

ipmatch [options] <images> <vwip files>

Command-line options for ipmatch:

--output-prefix <file name>

Write output files using this prefix.

--matcher-threshold <float (default: 0.6)>

Threshold for the separation between closest and next closest interest points.

--non-kdtree

Use a non-KDTree version of the matching algorithm.

--distance-metric <L2|Hamming (default: L2)>

Distance metric to use. Hamming should only be used for binary types like ORB.

--ransac-constraint <similarity|homography|fundamental|none>

RANSAC constraint type.

--inlier-threshold <float (default: 10)>

RANSAC inlier threshold.

--ransac-iterations <integer (default: 100)>

Number of RANSAC iterations.

-d, --debug-image

Set to write out debug images.

--threads <integer (default: 0)>

Select the number of threads to use for each process. If 0, use the value in ~/.vwrc.

--cache-size-mb <integer (default = 1024)>

Set the system cache size, in MB.

--tile-size <integer (default: 256 256)>

Image tile size used for multi-threaded processing.

--no-bigtiff

Tell GDAL to not create bigtiffs.

--tif-compress <None|LZW|Deflate|Packbits (default: LZW)>

TIFF compression method.

-v, --version

Display the version of software.

-h, --help

Display this help message.