Fig. 1

Edge detection pipeline \(\varvec{\mathcal{P}}\). The pipeline is split into five parts: (a) \(S\), the set of input images in RGB color space, (b) the function \({\mathcal{A}}\) that takes \(G\) as input and produces a boundary image set \(S\), (c) the output of \({\mathcal{A}}\) as gray scale images, (d) the edge extraction function \({\mathcal{E}}\) with parameters \(\{\alpha , \beta , \gamma \}\) that converts boundaries into edges, (e) the output \(O\) of the function \({\mathcal{E}}\), the set of binary images containing thin edges. Figure 3 shows a specific example this pipeline