Edge detection, gradient detection, fuzzy image

The contribution serves as a supporting report for outputs posted for EUSFLAT Competition on Edge Detection 2017. We present three different types of methods used for edge detection in an image. The methods differ in their interpretation of the term edge. The first one considers edges as thresholded gradient magnitudes. The second one reduces edges thickness in order to obtain 1px thin edges. The last one focuses on obtaining 1 pixel thin and continuous edges. The contribution describes the three methods, demonstrates their visual outputs and points their advantages and disadvantages.