Super Pixel Seg
Segment an image using super-pixels.
By Laurent Ittiitti@usc.eduhttp://jevois.orgGPL v3
 Language: C++Supports mappings with USB output: YesSupports mappings with NO USB output: No 
 Video Mapping:   GREY 320 240 30.0 YUYV 320 240 30.0 JeVois SuperPixelSeg
This module has no screenshots and no videos

Module Documentation

Segment an image into regions with somewhat uniform appearance, using the SuperPixel algorithms of OpenCV.

The result is an image with a reduced number of grey levels, where each grey level represents the label or ID of a given region (so, all pixels with a given grey value belong to a given region in the image).

How to use this module

When you run this algorithm and obtain a low number of clusters, those may not be well visible to the human eye. For example, if you end up with 10 clusters, they will take grayscale values 0 to 9, which all look completely black on a standard monitor where 0 is black and 255 is white. To get started with this module, you may hence want to change the parameters a bit. For example:

setpar algo SEEDS
setpar output Labels
setpar numpix 255

and you should now see something.

To work with a smaller number of super-pixels, you would usually want to first create some software that runs on the host computer and which will grab the greyscale frames from JeVois, then will assign colors to the regions somehow, and finally will display the superpixels in color. For testing, you may want to just capture and save some frames from JeVois (which may look all black) and then use some paint program to change color 0, color 1, etc to more visible colors than very similar shades of black.

ParameterTypeDescriptionDefaultValid Values
(SuperPixel) algoAlgoAlgorithm variant to useAlgo::SLICOAlgo_Values
(SuperPixel) regionsizeintSLIC Region size30-
(SuperPixel) numpixintNumber of SEEDS superpixels200-
(SuperPixel) iterationsintNumber of iterations6jevois::Range<int>(1, 100)
(SuperPixel) outputOutTypeOutput labels or demo grayscale input image with superimposed contoursOutType::ContoursOutType_Values
Detailed docs:SuperPixelSeg
Copyright:Copyright (C) 2016 by Laurent Itti, iLab and the University of Southern California
License:GPL v3
Support URL:
Other URL:
Address:University of Southern California, HNB-07A, 3641 Watt Way, Los Angeles, CA 90089-2520, USA