ARToolKit NFT Utilities: genTexData

From ARToolworks support library

Main Page > ARToolKit NFT > ARToolKit NFT Utilities: genTexData

Purpose

genTexData performs training of NFT datasets from a supplied JPEG-format source image.

Usage

./genTexData <filename>
    -with_imageSet2
    -level=n
         (n is an integer in range 0 (few) to 3 (many). Default 1.'
    -sd_thresh=<sd_thresh>
    -max_thresh=<max_thresh>
    -min_thresh=<min_thresh>
    -leveli=n
         (n is an integer in range 0 (few) to 3 (many). Default 1.'
    -surf_thresh=<surf_thresh>
    -dpi=<dpi>
    -max_dpi=<max_dpi>
    -min_dpi=<min_dpi>
    -background
         (Mac OS X and Linux only.)
    -log=<path>
    -exitcode=<path>

Operating example

Launch the program from a terminal window, supplying the name of the input JPEG file as the parameter.

You can specify how selective the training algorithms should be in rejecting candidate tracking features.

The utility will load the JPEG and read its size and expected printed resolution. If no resolution value is embedded, you will be prompted to supply the resolution.

Specify the minimum and maximum resolution values for tracking. A suitable image set will then be automatically generated from this range, and saved to disk with the suffix ".iset".

The utility begins to generate tracking data. This procedure may take some time, even on a fast CPU.

When generation of tracking data is complete, the utility will save the ".fset" and ".fset2" files to disk alongside the .jpeg. Training is then complete.