The calculation of land surface temperatures (LSTs) via low-altitude thermal infrared remote (TIR) sensing images at a block scale is gaining attention. However, the accurate calculation of LSTs requires a precise determination of the range of various underlying surfaces in the TIR images, and existing approaches face challenges in effectively segmenting the underlying surfaces in the TIR images. To address this challenge, this study proposes a deep learning (DL) methodology to complete the instance segmentation and quantification of underlying surfaces through the low-altitude TIR image dataset. Mask region-based convolutional neural networks were utilized for pixel-level classification and segmentation with an image dataset of 1350 annotated TIR images of an urban rail transit hub with a complex distribution of underlying surfaces. Subsequently, the hyper-parameters and architecture were optimized for the precise classification of the underlying surfaces. The algorithms were validated using 150 new TIR images, and four evaluation indictors demonstrated that the optimized algorithm outperformed the other algorithms. High-quality segmented masks of the underlying surfaces were generated, and the area of each instance was obtained by counting the true-positive pixels with values of 1. This research promotes the accurate calculation of LSTs based on the low-altitude TIR sensing images.

Recently, computer vision (CV)-based techniques have been widely applied for inspecting shield tunnel lining defects. However, this inspection method will produce a lot of image data. How to efficiently detect water leakage from the large image database therefore becomes a new challenge. To effectively improve recognition of water leakage of metro shield tunnel lining, a Mask Region-based Convolutional Neural Network (Mask R-CNN)-based method is employed in this paper. This approach efficiently detects water leakage in an image while simultaneously generating a high-quality segmentation mask for each water leakage. To realize the proposed method, a database including 4525 images is developed. Then, the Mask R-CNN is modified and trained using this database to detect and segment water leakage in images. The segmentation results show that the trained model shows quite better performances and can rapidly and accurately recognize water leakage of metro shield tunnels.