In order to promote evaluation of unsupervised color image segmentation algorithms using publicly available data sets, standard performance assessment methodology and on-line web verification server and database, a competition for the best segmentation algorithms will take place in conjunction with the ICPR 2014 conference.
Although numerous different methods were already published the ill-defined segmentation problem is still far from being solved. In addition to that, very little is known about properties and behaviour of already published segmentation methods and their potential user is left to randomly select one due to absence of any counselling.
The contest aim is to overcome these problems by suggesting the most promising approaches to the unsupervised learning and image segmentation and to unify the verification methodology used in the image segmentation research.
The authors of the best performing algorithm of the competition will receive a prize that will be presented at ICPR 2014. The performance of all submitted algorithms will be summarised in a presentation given at the conference. The best participants will be invited to publish their approach in a special journal issue.