Abstract:
Gradient-based edge detection is a straightforward method
to identify the edge points in the original grey-level image.
It is consistent with the intuition that in the human vision
system the edge points always appear where the change of
grey-level is greatest within their neighbourhood. In this
paper, triple-diagonal gradient-based edge detection is
introduced. It is based on the features of Spiral
Architecture and computes the gradients in three diagonal
directions instead of approximating the gradient in one
direction only as the traditional methods do. Essentially, it
improves the accuracy for locating edge points. As a
result, it does not need any supplementary processing to
enhance the edge map.
