Smooth curve and fitted curve-based road surface crack extraction algorithm

Abstract

The invention discloses a smooth curve and fitted curve-based road surface crack extraction algorithm. The algorithm specifically comprises the following steps of storing acquired road surface three-dimensional image data in an image data matrix AmXn, and first preprocessing the matrix AmXn by using a breakpoint interpolation algorithm with a gradually changed slope to obtain a processed data matrix A'mXn; then performing bidirectional standard deviation filtration processing on the preprocessed road surface data matrix A'mXn to obtain a filtered road surface three-dimensional image data matrix A''mXn; and finally extracting each row of data of the image data matrix A''mXn, performing smoothing and curve fitting processing separately, and extracting a contour line of a road surface crack according to a smooth curve and a fitted curve, thereby extracting the crack. The algorithm is low in complexity and short in runtime, and no artificial participation is required; and compared with an algorithm used for improving a basic line and proposed according to Fourier filtering, the extraction algorithm has the advantages that a reference line obtained by the extraction algorithm is continuous and smooth, so that the crack can be treated later more conveniently.

Description

technical field [0001] The invention belongs to the field of road engineering, in particular to a road crack extraction algorithm based on smooth curves and fitting curves. Background technique [0002] Cracks are one of the most common, most prone and earliest diseases among all kinds of pavement damage, and are generally divided into transverse cracks, longitudinal cracks and network cracks. The initial cracks have no obvious impact on the performance of the pavement, but they accompany the entire service life of the road. If they are not dealt with in time, the cracks will gradually increase under the influence of environmental factors and the repeated action of a large number of driving loads. Cracks on the pavement not only affect the appearance of the road and the comfort of driving, but also easily expand to cause structural damage to the pavement, shorten the service life of the pavement, and even affect the safety of driving to a large extent. Therefore, how to det...

AI Technical Summary

Problems solved by technology

[0005] In order to solve the problems of high detection cost, long time consumption and low accuracy of detection results of existing detection technologies and the improved baseline algorithm based on Fourier filtering in the prior art, the extracted road baseline is broken and cannot form a continuous The benchmark curve makes the extracted cracks not perfect and not accurate enough. The present invention adopts the following technical solutions to solve the problem:

Images