Control Point Detector Parameters

This site will give you an overview about the different control point detectors and their main options to use with hugin. This settings can be changed in preferences dialog. A detailed description of the individual settings can be found here

Cpfind

• For outdoor projects, single and multi-row panorama

Detector Type	All images at once
Detector	One-step detector
Program	cpfind
Arguments	--multirow --celeste -o %o %s

• For indoor projects, single and multi-row panorama

Detector Type	All images at once
Detector	One-step detector
Program	cpfind
Arguments	--multirow -o %o %s

• For prealigned panoramas (e.g. when using a template or script)

Detector Type	All images at once
Detector	One-step detector
Program	cpfind
Arguments	--prealigned -o %o %s

• For very big outdoor projects, single and multi-row panorama, with caching to disc

Detector Type	Multi-row panoramas
Detector	One-step detector
Program	cpfind
Arguments	--cache --celeste -o %o %s
Cleanup arguments	--clean %s

Hints:

• MacOSX: Both in the bundle version as in the cmake version you need to add cpfind without browsing for it and without adding a pathname, just cpfind.

Panomatic

• For general projects

Detector Type	All images at once or Multi-row panorama
Detector	One-step detector
Program	panomatic
Arguments	-o %o %i

• For single-row projects

Detector Type	All images at once
Detector	One-step detector
Program	panomatic
Arguments	--linearmatch -o %o %i

Autopano-sift-C

• For general projects

Detector Type	All images at once
Detector	One-step detector
Program	autopano-sift-c
Arguments	--maxmatches %p --projection %f,%v %o %i in version in Mercurial the parameter --projection is broken, use instead --maxmatches %p %o %i or --maxmatches %p %o %s (the latter works also for bigger projects)

• For multi-row projects

Detector Type	Multi-row panorama
Detector	Two-step detector
Feature descriptor	generatekeys
Arguments	%i %k 800
Feature matcher	autopano
Arguments	--maxmatches %p %o %k

Autopano by Alexandre Jenny

Detector Type	Autopano (by A. Jenny)
Detector	One-step detector
Program	autopano
Arguments	/allinone /path:%d /keys:%p /project:oto /name:%o /size:1024 /f %i

Hints:

• Autopano runs only on Windows and Linux. There is no MacOS version.
• Don't mix autopano by A. Jenny with autopano from autopano-sift-C package.

Align_image_stack

• As standalone detector for HDR images

Detector Type	All images at once
Detector	One-step detector
Program	align_image_stack
Arguments	-f %v -v -p %o %i or -f %v -v -e -p %o %i for full-frame fisheye images

Align_image_stack can also be used in conjugation with the above mentioned detector as special case for stacks. For this use case set Detector Type to Panorama with stacks or Multi-row panorama with stacks and set the options for align_image_stack on the tab Detector for stacks

Hints:

• MacOSX: Both in the bundle version as in the cmake version you need to add align_image_stack without browsing for it and without adding a pathname, just align-image_stack.

Linefind

• Vertical features autodetector

Detector Type	All images at once
Detector	One-step detector
Program	linefind
Arguments	-o %o %s

Geocpset

• Add geometric control points to images without features. This add control points based on the current position of the images.
• Usecase: projects containing images without features (e. g. images with only sky)
• Attention: You need to assign the position to the images before running this control point detector. Use the drag feature in the preview window, enter numeric values for yaw, pitch and roll in the Photos tab or use pto_var to assign the position. Running on a not aligned images will create wrong control points.

Detector Type	All images at once
Detector	One-step detector
Program	geocpset
Arguments	-o %o %s