EUSFLAT 2017 Competition on Edge Detection

Objectives

Detect edges in 30 images of the Dataset ED17 using any kind of detectors combined with methods of computational intelligence. The competition is conducted for pure academic purposes and supported by the EUSFLAT Working Group on Soft Computing Methods in Image Processing.

Methods

The Edge Detection competition is open to all methods of computational intelligence, incl. applications of aggregation operators, overlaps, groupings, etc., mathematical morphology, techniques of various transforms, convolutions and various methods of computational intelligence.

Publication of Results

We solicit participants to send their results to the 10th conference of the European Society for Fuzzy Logic and Technology -- EUSFLAT-2017, Warsaw, Poland, September 11-15. The conference will provide best paper & best submission diplomas for winners (students and non-students). All results are invited to be prepared in the form of contributions and submitted to the conference special session „Image Processing with Fuzzy Techniques“.

The results and winners will be announced on the EUSFLAT-2017 conference.

Instructions

1. Download the competition datasets.
2. Develop an original methodology for edge detection - in software code or documented steps.
3. Describe the methodology in a brief summary (up to 6 pages) using the EUSFLAT template for regular submissions.
4. Detect edges from images in the Dataset ED17. Edges have to be identified with pixels of white color (#FFFFFF), non-edges with pixels of black color (#000000) -- that is, the solution has to be a set of black/white images.
5. Save the results of edge detection in PNG files with names derived from original file names by replacing "d" with "s", e.g. the edges of the original file name d1-0001.png should be named s1-0001.png. Compress and record the results together with a PDF file describing your methodology in a ZIP data file with your name.
6. Store the ZIP file somewhere on the Internet, where it can be downloaded by organizers (to a cloud service of your choice - e.g. Dropbox).
7. Submit the link to your solution ZIP file via the form on this website to the specified deadline.

We look forward to your submission and also to meet you in Warsaw on EUSFLAT-2017 conference.

Kind regards & good luck!

Irina Perfilieva, Humberto Bustince, and Javier Montero