Scattered Train Bolt Point Cloud Segmentation Based on Hierarchical Multi-Scale Feature Learning

Abstract

In view of the difficulty of using raw 3D point clouds for component detection in the railway field, this paper designs a point cloud segmentation model based on deep learning together with a point cloud preprocessing mechanism. First, a special preprocessing algorithm is designed to resolve the problems of noise points, acquisition errors, and large data volume in the actual point cloud model of the bolt. The algorithm uses the point cloud adaptive weighted guided filtering for noise smoothing according to the noise characteristics. Then retaining the key points of the point cloud, this algorithm uses the octree to partition the point cloud and carries out iterative farthest point sampling in each partition for obtaining the standard point cloud model. The standard point cloud model is then subjected to hierarchical multi-scale feature extraction to obtain global features, which are combined with local features through a self-attention mechanism, while linear interpolation is used to further expand the perceptual field of local features of the model as a basis for segmentation, and finally the segmentation is completed. Experiments show that the proposed algorithm could deal with the scattered bolt point cloud well, realize the segmentation of train bolt and background, and could achieve high segmentation accuracy, which has important practical significance for train safety detection.

Soil Modeling and Prediction Methods in Dredging Construction Areas

“…Moreover, Cao et al [4] established a Bayesian framework for probabilistic soil stratification to identify soil, while Wise et al [5] proposed the use of a combination of 3D points and acoustics to predict soil erosion, aimed at three-dimensional point clouds. For noisy points in point cloud models, Zeng et al [6] used a point cloud adaptive weighted guided filtering algorithm to smooth out noise based on its characteristics, which can effectively preserve the key points of the point cloud. In order to accurately predict the local details of point clouds, Hao et al [7] proposed a new adaptive point cloud growth grid, MapGNET, to generate higher-quality point clouds.…”

A cloud model-based probabilistic linguistic multi-attribute decision method and its application to public evacuation in nuclear accidents

“…Cloud model is an uncertainty transformation model between a qualitative concept and its quantitative representation using linguistic values, which mainly reflects the uncertainty of things in the objective world or concepts in human knowledge, and constitutes a mutual mapping between qualitative and quantitative. Since Li et al (2009) proposed the cloud model, it has been widely used in many fields such as intelligent control (Ma et al 2023;Zeng et al 2023), risk assessment (Zou et al 2023Yu et al 2023), system measurement (Yang et al 2023;Ye et al 2023;) and decision making. In the research in the field of decision making, aggregated fuzzy decision information in the form of cloud models to obtain decision results, and proposed a large-scale group classification decision method with dual trust-benefit factors in social networks.…”