Vector extraction from digital images of steel bars produced by an inductive scanning system using a differential gradient method combined with a modified Hough transform

A method of extracting positional information from images of steel bars embedded in concrete using a set of image pre-processing algorithms combined with a modified Hough transform is described. The images are formed using a portable inductive scanning system in which a sensing element detects the magnetic field emanating from the bars. Eddy currents induced on the surfaces of the bars by excitation coils generate these magnetic fields. The results from an edge detection algorithm based on the two-dimensional separable smoothing and local maxima detection are applied to a two-stage Hough transform which locates the major axes of the bars and hence their alignment and number. An effective method of minimizing the redundant information in the Hough array is also described. The complete system incorporates fully integrated control, acquisition and processing software, which makes it convenient for on-site use.