METHOD FOR EXTRACTING SURFACE CRACKS OF PAVED ROAD AND EVALUATING SURFACE DAMAGE LEVEL OF PAVED ROAD

摘要

PROBLEM TO BE SOLVED: To solve a problem that the photographed images of spots are carried back to an inspection station in the detection of cracks of a paved road to be analyzed and evaluated using an analyzing device but quantities of the processed images are large in the detection of the cracks of a road of which the total length reaches a several-hundred kilometer or longer and unsuitable for man-power work, and since evaluation also relies the subjective determination by the experience of a technical person, it is impossible to subjectively and accurately perform the analysis and evaluation of the cracks of the paved road.;SOLUTION: In this method for extracting the surface cracks of the paved road and evaluating the damage level of the paved road, size-dependent crossing is introduced into a parallel image filter automatic forming system by genetic programming to select and employ places considered to be difficult in the extraction of the crack (1) from a plurality of actual pavement images in the training data of filter construction to thereby automatically construct the extraction image filter (4) of the crack from images of various of types. Then, the extraction image filter (4) is adapted to the individual blocks (6) formed by dividing the whole image of a pavement region to be evaluated in a lattice shape to perform the evaluation of the same.;COPYRIGHT: (C)2011,JPO&INPIT