Universitaet Koblenz-Landau

Universitaet Koblenz-Landau Fachbereich Informatik
Image Recognition Laboratory
Image Recognition Laboratory Download Page
Ansprechpartner (Email, Telefon)
Dokumentenbestand dieses Servers durchsuchen
Informationein und Hilfe zu diesem Server
 

Manual: Color Structure Code (CSC)

The Algorithm

The CSC is a hierarchical region growing method. It detects homogenous color regions in images. Further information about the Color Structur are available in the online lecture (lectures.ps,lectures.pdf ).


Description

The zip archives csc-linux-1-0.zip and csc-windows-1-0.zip contain each the binary version of the command line tool csc. With the command line tool csc you are able to make segmentations of color images with the color structure code. The 'csc' needs as input a color image in PPM-Format (Portable Pixmap in binary format with magic number 'P6'). It produces two images: a region image that stores for each pixel the average color of its region and a labeled image that stores for each pixel a label number of its region.

Usage

Syntax: csc [options] "name_of_input_image" "name_of_region_image" "name_of_labeled_image"

Options Reference:

-c [rgb/hsv]: This options defines on which color space the color similarity predicate should operate.
-t threshold:Threshold is an integer value that defines the smallest allowed euclidian distance between two similar rgb color vectors The default value of threshold is 16. If you choose a small threshold the csc will produce more segments than with a large threshold. (This option works only in combination with -c rgb).
-h hue-table:With hue-table [-4..4] you can select one of 9 distance tables containing hue thresholds. If you choose a table with a high number, less color nuances could be distinguished. Therefore the csc will produce less segments than by using a small "hue-table number". The default value is 0. (This option works only in combination with -c hsv)
-s sat-table:With sat-table [-4..4] you can select one of 9 distance tables containing saturation thresholds. If you choose a table with a high number, less saturation nuances could be distinguished. Therefore the csc will produce less segments than by using a small "sat-table number". The default value is 0. (This option works only in combination with -c hsv)
-v val-table:With val-table [-4..4] you can select one of 9 distance tables containing val thresholds. If you choose a table with a high number, less value nuances could be distinguished. Therefore the csc will produce less segments than by using a small "val-table number". The default value is 0. (This option works only in combination with -c hsv)
Remarks: For the hsv color similarity predicate exists 9x9x9 degrees of freedom.

Example:

Suppose you want to make csc segmentation of an color image with name example.ppm. First be sure that your color image is in PPM graphics format and the PPM file has the magic number P6. The command
csc -c hsv -h 0 -s 2 -h 4 example.ppm region.ppm
will produce a region image with name region.ppm in the ppm format.




Editor: Patrick Sturm
Letzte Änderung: 15.07.2002