Search

Convolutional neural networks (CNNs) and Transformers have shown advanced accuracy in crack detection under certain conditions. Yet, the fixed local attention can compromise the generalisation of CNNs, and the quadratic complexity of the global self-attention restricts the practical deployment of Transformers. Given the emergence of the new-generation architecture of Mamba, this paper proposes a Vision Mamba (VMamba)-based framework for crack segmentation on concrete, asphalt, and masonry surfaces, with high accuracy, generalisation, and less computational complexity. Having 15.6% - 74.5% fewer parameters, the encoder-decoder network integrated with VMamba could obtain up to 2.8% higher mDS than representative CNN-based models while showing about the same performance as Transformer-based models. Moreover, the VMamba-based encoder-decoder network could process high-resolution image input with up to 90.6% lower floating-point operations., Comment: 23 pages, 9 figures

Vision-based crack detection faces deployment challenges due to the size of robust models and edge device limitations. These can be addressed with lightweight models trained with knowledge distillation (KD). However, state-of-the-art (SOTA) KD methods compromise anti-noise robustness. This paper develops Robust Feature Knowledge Distillation (RFKD), a framework to improve robustness while retaining the precision of light models for crack segmentation. RFKD distils knowledge from a teacher model's logit layers and intermediate feature maps while leveraging mixed clean and noisy images to transfer robust patterns to the student model, improving its precision, generalisation, and anti-noise performance. To validate the proposed RFKD, a lightweight crack segmentation model, PoolingCrack Tiny (PCT), with only 0.5 M parameters, is also designed and used as the student to run the framework. The results show a significant enhancement in noisy images, with RFKD reaching a 62% enhanced mean Dice score (mDS) compared to SOTA KD methods., Comment: 24 pages, 13 figures

Dynamic wetting plays an important role in the physics of multiphase flow, and has significant influence on many industrial and geotechnical applications. In this work, a modified smoothed particle hydrodynamics (SPH) model is employed to simulate surface tension, contact angle, and dynamic wetting effects. The wetting and dewetting phenomena are simulated in a capillary tube, where the liquid particles are raised or withdrawn by a shifting substrate. The SPH model is modified by introducing a newly-developed viscous force formulation at liquid-solid interface to reproduce the rate-dependent behaviour of moving contact line. Dynamic contact angle simulations with interfacial viscous force are conducted to verify the effectiveness and accuracy of this new formulation. In addition, the influence of interfacial viscous force with different magnitude on contact angle dynamics is examined by empirical power law correlations, and the derived constants suggest the dynamic contact angle changes monotonically with interfacial viscous force. The simulation results are consistent with the experimental observations and theoretical predictions, implying that the interfacial viscous force can be associated with slip length of flow and microscopic surface roughness. This work has demonstrated that the modified SPH model can successfully account for the rate-dependent effects of moving contact line, and can be used for realistic multiphase flow simulation under dynamic conditions., Comment: 23 pages, 13 figures