SLIDE: subspace-based line detection

The SLIDE (subspace-based line detection) algorithm, a technique for estimating parameters of multiple straight lines in an image, is described. By reformulating the line fitting problem into a spectral estimation framework, SLID exploits subspace-based techniques of sensor array processing to obtain high resolution and closed-form estimates for the line parameters. The computational complexity of SLIDE is an order of magnitude less than that of the Hough transform method, and, unlike the Hough transform, SLIDE does not require a search procedure to estimate the parameters. Potential application areas of this technique include road tracing in robotic vision, aerial image analysis, mask-wafer alignment and linewidth measurement in semiconductor manufacturing, and text alignment in document analysis.<
>