Search

Digital camera was used to record the dynamic process of soil desiccation cracks in real time. With application of image processing technique, the relationship between the strain/displacement field and desiccation cracking were analyzed. Results show that the crack usually occurs at locations where tensile strain concentrates. After the crack is generated, the surrounding strain field and displacement field are rapidly redistributed. The new strain field and displacement field act on the crack again, affecting the propagation direction of the crack. Under the interaction between the strain/displacement field and cracks, the crack tends to intersect with other cracks at right angles and deflection may occur in the development process. The soil is divided into different clods by the cracks. Each clod centripetally shrinks and has a shrinkage center which may move with elapsed drying time. Based on the strain/displacement field, the characteristics of desiccation cracking in local area of soil can be predicted in advance and it can be used to directly judge the activeness of the cracks or clods. The digital image processing technique can effectively obtain the soil deformation without disturbing the soil sample, which offers an available way for understanding the intrinsic dynamic mechanisms of soil desiccation cracking. [ABSTRACT FROM AUTHOR]

Drought-induced soil desiccation cracking has attracted great attention in various disciplines with the advent of global climate change. Accurately obtaining soil crack networks is essential to understand the cracking mechanism. Inspired by recent advances of artificial intelligence (AI) in computer vision, we propose a new automatic soil crack recognition method based a novel network architecture, named Attention Res-UNet. Deep Res-UNet inherits both the advantages from residual learning for training deeper networks and U-Net for semantic segmentation. Moreover, attention mechanism is utilized to alleviate the influences caused by the uneven illumination conditions. Firstly, the soil crack images under different uneven illumination conditions are collected to create a new soil cracking image dataset. Then, traditional method and multiple state-of-the-art deep learning based different semantic segmentation models are tested on our collected dataset. Finally, a professional evaluation standard, which considers both the overall metrics (precision, recall, dice, surface crack ratio) and details (crack total length, average crack width, number of crack segments) of the soil crack features is proposed to evaluate the recognition results of the different models. Extensive experimental results demonstrate the superiority of our proposed Attention Res-UNet approach compared with traditional methods and other deep learning models in recognizing soil cracks under complex environmental conditions. Our method is also suitable for crack recognition of other materials under complex environmental conditions. • Artificial intelligence is used for soil cracks recognition. • Deep Res-UNet and Attention gate is combined for soil crack image recognition. • A professional evaluation standard is proposed to evaluate recognition performance. • Proposed model is superior to other deep learning models in soil cracks recognition. [ABSTRACT FROM AUTHOR]