Most natural disasters result from geodynamic events such as landslides and slope collapse. These failures cause catastrophes that directly impact the environment and cause financial and human losses. Visual inspection is the primary method for detecting failures in geotechnical structures, but on-site visits can be risky due to unstable soil. In addition, the body design and hostile and remote installation conditions make monitoring these structures inviable. When a fast and secure evaluation is required, analysis by computational methods becomes feasible. In this study, a convolutional neural network (CNN) approach to computer vision is applied to identify defects in the surface of geotechnical structures aided by unmanned aerial vehicle (UAV) and mobile devices, aiming to reduce the reliance on human-led on-site inspections. However, studies in computer vision algorithms still need to be explored in this field due to particularities of geotechnical engineering, such as limited public datasets and redundant images. Thus, this study obtained images of surface failure indicators from slopes near a Brazilian national road, assisted by UAV and mobile devices. We then proposed a custom CNN and low complexity model architecture to build a binary classifier image-aided to detect faults in geotechnical surfaces. The model achieved a satisfactory average accuracy rate of 94.26%. An AUC metric score of 0.99 from the receiver operator characteristic (ROC) curve and matrix confusion with a testing dataset show satisfactory results. The results suggest that the capability of the model to distinguish between the classes 'damage' and 'intact' is excellent. It enables the identification of failure indicators. Early failure indicator detection on the surface of slopes can facilitate proper maintenance and alarms and prevent disasters, as the integrity of the soil directly affects the structures built around and above it.

In response to increasing Arctic temperatures, ice-rich permafrost landscapes are undergoing rapid changes. In permafrost lowlands, polygonal ice wedges are especially prone to degradation. Melting of ice wedges results in deepening troughs and the transition from low-centered to high-centered ice-wedge polygons. This process has important implications for surface hydrology, as the connectivity of such troughs determines the rate of drainage for these lowland landscapes. In this study, we present a comprehensive, modular, and highly automated workflow to extract, to represent, and to analyze remotely sensed ice-wedge polygonal trough networks as a graph (i.e., network structure). With computer vision methods, we efficiently extract the trough locations as well as their geomorphometric information on trough depth and width from high-resolution digital elevation models and link these data within the graph. Further, we present and discuss the benefits of graph analysis algorithms for characterizing the erosional development of such thaw-affected landscapes. Based on our graph analysis, we show how thaw subsidence has progressed between 2009 and 2019 following burning at the Anaktuvuk River fire scar in northern Alaska, USA. We observed a considerable increase in the number of discernible troughs within the study area, while simultaneously the number of disconnected networks decreased from 54 small networks in 2009 to only six considerably larger disconnected networks in 2019. On average, the width of the troughs has increased by 13.86%, while the average depth has slightly decreased by 10.31%. Overall, our new automated approach allows for monitoring ice-wedge dynamics in unprecedented spatial detail, while simultaneously reducing the data to quantifiable geometric measures and spatial relationships.