cpclean - Man Page
Remove wrong control points by statistical methods
Synopsis
cpclean [options] input.pto
Description
cpclean uses statistical methods to remove wrong control points.
Step 1 optimises all images pairs, calculates for each pair mean and standard deviation and removes all control points with error bigger than mean+n*sigma.
Step 2 optimises the whole panorama, calculates mean and standard deviation for all control points and removes all control points with error bigger than mean+n*sigma.
Options
- --output|-o output.pto
Output Hugin PTO file. Default: '<filename>_clean.pto'.
- --max-distance|-n num
distance factor for checking (default: 2) (cps with an error > mean + this factor*sigma will be removed)
- --pairwise-checking|-p
do only image pair cp checking (skip step 2)
- --whole-pano-checking|-w
do only whole panorama checking (skip step 1)
- --dont-optimize|-s
skip optimisation step when checking the whole panorama
- --check-line-cp|-l
also include line control points for calculation and filtering in step 2
- --verbose|-v
verbose output during optimisation
- --help|-h
shows help
Authors
Thomas Modes