This paper does not report on the accuracy of the techniques used. A three-dimensional model of the structure is also created to better contextualise these defects. The solution presented also implements a change detection algorithm to attract attention to areas that are considered important, and the inspectors can refer to previous imagery from previous inspections to assess the evolution of the defect. An array of cameras with uniform lighting is used to capture the image data that are registered and stitched to allow tunnel inspectors to examine for defects with reference to their location. A system aimed at supporting structural inspectors to monitor the condition of railway tunnels is presented in. However, literature on the detection of changes on tunnel linings is still lacking, possibly due to the challenges encountered in this area. Reviews of change identification methods are found in. Furthermore, the outcome from these inspections is highly dependent on human subjectivity, leading to possible inaccuracies that result in false and missing change detections.Ĭhange detection is a well-researched problem in the fields of video surveillance, remote sensing () and medical imaging, amongst others. In addition, these inspections require a considerable amount of time to perform, leading to longer operation down-times and thus higher monetary losses. Associated with this, there are several drawbacks, including the human presence in hazardous environments and the financial cost involved to train and hire people to do the inspections. To conduct such observations, personnel are required to be physically present in the tunnel. Strain gauges, displacement meters and other contact measurement methods can be employed to monitor problematic areas more closely within these structures. Today, these are predominantly performed through periodic visual observations, looking for structural defects such as cracking, spalling and water leakage to identify possible changes in the infrastructure with respect to a previous survey. Consequently, periodic inspections of concrete tunnels should be conducted to ensure that they are still healthy and safe. Tunnel infrastructure shows signs of deterioration over time due to construction defects, ageing, unexpected overloading and natural phenomena, possibly leading to problems in structural integrity. A quantitative analysis of the results achieved shows that the proposed change detection system achieved a recall value of 81%, a precision value of 93% and an F1-score of 86.7%. Decision-level fusion methods are then used to combine these change maps to obtain a more reliable detection of the changes that occur between surveys. Different pixel-based change detection approaches are used to generate temporal change maps. Image fusion techniques are then applied to identify the changes occurring in the tunnel structure. These data are then pre-processed using image processing and deep learning techniques to reduce nuisance changes caused by light variations. The vision-based sensing system acquires the data from a rig of cameras hosted on a robotic platform that is driven parallel to the tunnel walls. In this work, we propose a remotely operated machine vision change detection application to improve the structural health monitoring of tunnels. These issues can be mitigated through accurate automatic monitoring and inspection systems. These surveys are subjective (relying on the surveyor experience), time-consuming, and may demand operation shutdown. Tunnel structural health inspections are predominantly done through periodic visual observations, requiring humans to be physically present on-site, possibly exposing them to hazardous environments.
0 kommentar(er)
