Extracting full-field subpixel structural displacements from videos via deep learning
This work addresses the need for real-time, accurate displacement monitoring in structural engineering, though it is incremental as it builds on existing phase-based motion extraction methods.
The paper tackles the problem of extracting full-field subpixel structural displacements from videos by developing a deep learning framework based on CNNs, achieving real-time capability and generalizability to accurately extract subtle displacements for pixels with sufficient texture contrast.
This paper develops a deep learning framework based on convolutional neural networks (CNNs) that enable real-time extraction of full-field subpixel structural displacements from videos. In particular, two new CNN architectures are designed and trained on a dataset generated by the phase-based motion extraction method from a single lab-recorded high-speed video of a dynamic structure. As displacement is only reliable in the regions with sufficient texture contrast, the sparsity of motion field induced by the texture mask is considered via the network architecture design and loss function definition. Results show that, with the supervision of full and sparse motion field, the trained network is capable of identifying the pixels with sufficient texture contrast as well as their subpixel motions. The performance of the trained networks is tested on various videos of other structures to extract the full-field motion (e.g., displacement time histories), which indicates that the trained networks have generalizability to accurately extract full-field subtle displacements for pixels with sufficient texture contrast.