A Computational Approach to Edge Detection

Detect thin step edges in greyscale images by smoothing with a Gaussian, computing gradient magnitude and direction, suppressing non-maxima along the gradient direction, then linking surviving pixels via hysteresis double-thresholding; the filter shape is derived as the variational optimum of three criteria — detection SNR, localisation, and single-response spacing — under an additive-white-Gaussian-noise step-edge model.

The linear, shift-invariant operation that produces each output pixel as a kernel-weighted sum of input pixels in a local neighbourhood.

Reducing a dense response map or a set of overlapping detections to a sparse set of local maxima by discarding every element that is not strongest in its neighbourhood.