Building crack monitoring method and system based on machine learning

Abstract

The invention discloses a building crack monitoring method and system based on machine learning, and relates to the technical field of building detection, and the method comprises the following steps: S1, multi-band image data enhancement and preprocessing; s2, carrying out self-adaptive segmentation on a crack region; s3, crack feature vectorization representation is carried out; s4, crack feature learning based on self-supervised learning; s5, crack generation based on an image generative adversarial network (GAN); s6, performing fracture morphological dynamics modeling; and S7, fracture topology analysis based on a graph structure. According to the building crack monitoring method and system based on machine learning, through self-adaptive segmentation, self-supervised learning and crack morphological dynamics modeling, the precision and robustness of crack monitoring are effectively improved, the problems of false detection and missing detection in crack detection of a traditional method are solved, and the detection accuracy is improved. The system can adaptively learn representation forms of cracks in different environments and materials, accurately extract crack features and predict expansion paths of the crack features.

Images