What are the advantages and disadvantages of various border detection algorithms
From an application point of view:
Sobel: Fast but not detecting faint edges, copes with noise
FuzzyLogic: Comparatively slow. Require tuning in the presence of noise (not sure how tedious)
Canny: Still considered good thanks to the non-maximum suppression and two thresholds that extract most of the ribs. Failure of bad in the presence of noise, manual parameter adjustment is required.
Search: Detect an edge from a google researcher and you get a detailed comparison.
source to share