KYKai Yang

Kai Yang

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Kai Yang has a total of 40 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Physics-guided label-free ultrasonic testing of high-speed rail wheels with diffusion models

    by Qian Zhang, Fuben Zhang, Ai Wang, Jinlong Li, Xiaorong Gao, Jianping Peng, Kai Yang

    1. The safety of high-speed rail (HSR) systems is of paramount importance, as failures in critical components can have severe consequences for the overall operation of the railway network [1]. Amon...

  • Single-shot 3D measurement with global phase offset correction based on stereo color phase-shifting profilometry

    by Tao Tang, Yu Zhang, Kai Yang, Yingying Wan, Jianping Peng, Lin Luo

    Phase-shifting profilometry (PSP) plays a dominant role in the field of three-dimensional (3D) surface-shape measurement due to its high accessible measurement accuracy and spatial resolution. However, PSP relies on multiple projections, which makes it prone to motion-induced errors and thus its performance is limited in dynamic scenes. In this paper, we introduce a novel single-shot 3D measurement method based on stereo color PSP. This method requires no additional images, benefiting from geometric constraints and colored phase-shifted fringes. The phase errors caused by color crosstalk and gamma nonlinearity are corrected by the chord distribution equalization method and the cross-correlation-based global phase offset correction method. The point cloud obtained using the corrected phase is reprojected onto the second camera plane, and the corresponding matching points are found on the epipolar lines within a limited search range. After stereo correspondence, the final 3D point cloud of the measured object is reconstructed based on stereo technology. Three dynamic scenes (a changing palm, a rotating Beethoven statue on a platform, and a rotating electric fan) are assessed to verify the robustness and accuracy of the proposed technique in a range of dynamic and complex environments.


  • Open Access

    Robust Line-Scan Image Registration via Disparity Estimation for Train Fault Diagnosis

    by Darui Feng, Kai Yang, Zhi Ling, Yong Wang, Lin Luo

    Automatic fault detection based on machine vision technology is crucial for the operational safety of trains. However, when imaging moving trains, system errors may induce localized geometric distortions in the captured images, altering the shapes of critical train components. This, in turn, undermines the precision of subsequent diagnostic algorithms. Therefore, image registration prior to anomaly detection is essential. To address this need, we redefine the horizontal registration of line-scan images as a disparity estimation problem on rectified stereo pairs, which is solved using a proposed dense matching network. The disparity is iteratively refined through a GRU-based update module that constructs a multi-scale cost volume with positional encoding and self-attention. To overcome the absence of real-world disparity ground truth, we generate a physics-based simulation dataset by analytically modeling the nonlinear relationship between train velocity variations and line-scan image distortions. Extensive experiments on diverse real-world train image datasets under varied operational conditions demonstrate that our method consistently outperforms alternatives, achieving 5.8% higher registration accuracy and a fourfold increase in processing speed over state-of-the-art approaches. This advantage is particularly evident in challenging scenarios involving repetitive patterns or texture-less regions.


  • Open Access

    A non‐rigid point cloud registration method based on scene flow estimation

    by Xiaopeng Deng, Kai Yang, Yong Wang, Jinlong Li, Liming Xie

    Point cloud registration is an important aspect of computer vision, encompassing various applications. However, many existing algorithms neglect the registration of non‐rigid point cloud. Recent studies have focused on scene flow, which can achieve non‐rigid point cloud registration by estimating the scene flow of two point clouds. The presence of occlusion in a scene is primarily attributed to variations in viewpoint and access time, thereby posing a significant challenge in accurately predicting scene flow. In the endeavour to mitigate this issue, the focus is on the propagation of scene flow originating from non‐occluded points towards occlusion points while concurrently estimating the occlusion map. The proposed network, a non‐rigid point cloud registration method based on scene flow estimation, achieves exceptional performance for the EPE3D metric on the FlyingThings3D and KITTI scene flow datasets, and it demonstrates strong generalization on the railroad dataset.


  • Open Access

    A Registration Method Based on Ordered Point Clouds for Key Components of Trains

    by Kai Yang, Xiaopeng Deng, Zijian Bai, Yingying Wan, Liming Xie, Ni Zeng

    Point cloud registration is pivotal across various applications, yet traditional methods rely on unordered point clouds, leading to significant challenges in terms of computational complexity and feature richness. These methods often use k-nearest neighbors (KNN) or neighborhood ball queries to access local neighborhood information, which is not only computationally intensive but also confines the analysis within the object’s boundary, making it difficult to determine if points are precisely on the boundary using local features alone. This indicates a lack of sufficient local feature richness. In this paper, we propose a novel registration strategy utilizing ordered point clouds, which are now obtainable through advanced depth cameras, 3D sensors, and structured light-based 3D reconstruction. Our approach eliminates the need for computationally expensive KNN queries by leveraging the inherent ordering of points, significantly reducing processing time; extracts local features by utilizing 2D coordinates, providing richer features compared to traditional methods, which are constrained by object boundaries; compares feature similarity between two point clouds without keypoint extraction, enhancing efficiency and accuracy; and integrates image feature-matching techniques, leveraging the coordinate correspondence between 2D images and 3D-ordered point clouds. Experiments on both synthetic and real-world datasets, including indoor and industrial environments, demonstrate that our algorithm achieves an optimal balance between registration accuracy and efficiency, with registration times consistently under one second.

  • Open Access

    Three-dimensional reconstruction based on binocular structured light with an error point filtering strategy

    by Xinyu Li, Kai Yang, Yingying Wan, Zijian Bai, Yunxuan Liu, Yong Wang, Liming Xie

    Optical 3D measurement technology has the advantages of being noncontact, having high precision and fast speed, and being applicable to static and dynamic measurements; it is widely used in fields such as machine vision, industrial inspection, biometrics, and reverse engineering.1 – 4 Binocular vision and structured light are important methods of optical 3D measurement technology.


  • Open Access

    A Robust End-to-End Speckle Stereo Matching Network for Industrial Scenes

    by Yunxuan Liu, Kai Yang, Xinyu Li, Zijian Bai, Yingying Wan, Liming Xie

    The detection capability of deep learning-based stereo matching in industrial applications is inherently limited due to challenges posed by weak texture and inconsistent reflectance, making it difficult to accurately recover complex surface details. To achieve accurate measurements, this paper presents an end-to-end speckle stereo matching network that incorporates fringe, Gray code, and speckle projection patterns. The model is trained using a high-precision dataset consisting of thousands of pairs generated through binocular Gray code-assisted phase shifting. After establishing local correspondences between the left and right images using speckle patterns, the images are used as inputs to the network. The proposed network consists of two siamese 2D feature extraction networks. One network is dedicated to cost volume computation, while the other focuses on weight refinement feature extraction. The former network incorporates a lightweight module for extracting high-dimensional fusion features. These features are obtained from different dilation scales and randomly concatenated along the channel dimension. Patch convolution is utilized to effectively adapt to pixel features at various levels, reducing redundancy within the cost volume and improving the network’s capacity to learn from ill-posed regions. Experimental results demonstrate that the proposed network achieves a significant improvement of approximately 10.7% in matching accuracy compared to state-of-the-art networks on public datasets. Furthermore, this method exhibits outstanding matching results when applied to diverse industrial scenarios. The reconstruction error for the radius of optical standard spheres is below 0.06-mm, which meets the demands of the majority of industrial applications.


  • An End-to-End Point-Based Method and a New Dataset for Street-Level Point Cloud Change Detection

    by Zhixue Wang, Yu Zhang, Lin Luo, Kai Yang, Liming Xie

    Studies on three-dimension change detection (3DCD) become a research hotspot with the development of 3D sensors. However, most of 3D change detection works are focused on remote sensing data. In the field of street level 3D data, the related works are under investigation. The two main challenges are the lack of point-wise annotated datasets and universal detection framework. In this paper, we proposed an end-to-end point-based network named 3DCDNet and a new dataset named street level point cloud change detection (SLPCCD) dataset to deliver street level 3DCD task. To structure the proposed 3DCDNet, a local feature aggregation (LFA) module and a nearest feature difference module (NFD) are introduced. The LFA is capable of extracting point features and aggregating local information, which is an effective neural module for embedding point cloud features. By stacking multiple of LFA blocks, the proposed network can be good at establishing relationships between different points and embedding semantically rich features. Different from the change detection in images, another crucial point in 3DCD is how to identify changes since point clouds are unstructured data. In order to deliver this challenge, the NFD module is introduced to identify change results using nearest query operation. Extensive experiments were implemented on the manually annotated SLPCCD dataset as well as another benchmark called Urb3DCD to validate the effectiveness and efficiency of the introduced network. It demonstrates that the proposed network outperforms popular existing approaches. The source code and the annotated dataset are available at: https://github.com/wangle53/3DCDNet.