Wrapper module that performs graph-based image segmentation exploiting the algorithm developed by Felzenszwalb and Huttenlocher (Brown University).

Description

Wrapper module that performs graph-based image segmentation exploiting the algorithm developed by Felzenszwalb and Huttenlocher (Brown University). The algorithm is described in the paper "Efficient Graph-Based Image Segmentation", Pedro F. Felzenszwalb and Daniel P. Huttenlocher, International Journal of Computer Vision, Volume 59, Number 2, September 2004. The original code is available from http://www.cs.brown.edu/~pff/segment/ and has been copied into the src/segmentation/graphBased/segment folder.

Libraries

YARP library.

Parameters

name name of the module (for opened ports)
sigma smoothing parameter (standard deviation of a Gaussian filter used to pre-process image, in order to compensate for digitization artifacts)
k scale factor for boundary-detection threshold function (a larger k causes a preference for larger components).
minRegion minumum size of a component.

Ports Accessed

None.

Ports Created

/<module_name>/rawImg:i input port where original images must be sent (default: GBSeg)
/<module_name>/viewImg:o
/<module_name>/conf rpc port, accepts the commands defined in graphBasedSegmentationInterface.thrift and documented here

Input Data Files

None

Output Data Files

None

Configuration Files

No configuration file is needed. Parameters described in Parameters can be provided in a configuration file; default filename is

Tested OS

Linux and Windows.

Example Instantiation of the Module

Examples of module use are provided in the app/scripts folder, and installed into app/graphBasedSegm/scripts.

Author: Elena Ceseracciu

CopyPolicy: Released under the terms of the GNU GPL v2.0 and later.

This file can be edited at src/segmentation/graphBased/main.cpp.