基于骨架和分形的混凝土裂缝图像识别算法

Skeleton and fractal law based image recognition algorithm for concrete crack

裂缝严重影响混凝土结构的安全性,为此研究人员开发各种检测技术来实现对裂缝的检测。基于数字图像的检测方法以其非接触、高精度等特点在实际工程中得到广泛地应用。分形理论属于非线性科学,适合用来描述无序的混乱现象和不规则的形态。利用摄像机拍摄清晰、完整的混凝土裂缝图像,结合数字图像处理技术和分形理论研究图像中裂缝及其他元素(诸如孔洞、杂质等)的自相似特征。通过分析混凝土裂缝图像面积-周长关系的分形规律,提出了一种结合简单灰度骨架提取和分形优化的裂缝识别算法。实验结果表明:裂缝与其他元素在分形特征上有着显著的不同,可利用该规律为混凝土表面裂缝的识别提供可靠的依据。

The cracks have a strong impact on the safety of concrete structures, and therefore engineers have devel oped a variety of detection techniques to achieve the recognition of cracks. The test method based on digital image, which features noncontact, highprecision, etc. , has been widely used in practical engineering fields. Fractal theo ry, which belongs to nonlinear science, is suitable to describe the chaotic confusion and irregular shape. In this pa per, camera is used to capture sharp and complete image of concrete cracks, and digital image processing technique and fractal theory are combined to study the selfsimilarity features of cracks and other elements （ such as holes, im purities, etc. ） in the image. Through analyzing the fractal law of area perimeter relation in concrete crack images, a recognition algorithm based on simple grayscale skeleton extraction and fractal optimization is proposed. Experi mental results show that cracks and other elements have significant difference in fractal features, which can provide a reliable method and basis for the recognition of the cracks on concrete surface.