We classify all the different kinds of errors that can occur in edge detection and then develop measures for quantifying these errors. It is shown that these sets of measures are complete and independent and form necessary components of an edge-evaluation scheme. The principle that an edge-evaluation measure should have certain qualitative properties is used to develop a method for combining these error components into a single combined measure.
Field of Research
080106 Image Processing
Socio Economic Objective
890205 Information Processing Services (incl. Data Entry and Capture)