Increasing numbers of people have taken up skiing in recent years due to the strong promotion of the Beijing Winter Olympics in 2022 by We Media. Nowadays, the establishment of three-dimensional digital twin snow fields has become an important way to effectively manage and maintain snow fields. However, due to the large altitude difference and the presence of many trees and rocks in ski resorts, traditional methods face difficulties such as low segmentation accuracy and low merging efficiency of segmentation blocks when performing semantic segmentation of ski runs. Consequently, this paper proposes a contour extraction method of artificial ski runs using composite supervoxels based on the characteristics of artificial ski resorts. To begin with, the point cloud data set of ski resort is segmented to get supervoxels; secondly, the difference in elevation between the seed supervoxel and the adjacent connecting block is calculated to determine whether the merging plane is the ground or another plane; then, according to the normal vector angle threshold and the orthogonal distance threshold, the similarity between the current clustering region and adjacent blocks is evaluated; and finally, the region growth algorithm is optimized based on the point cloud supervoxels of ski resorts, in order to reap the benefits of ski track semantic segmentation. And experiments have shown that the proposed method is superior to the other two in terms of segmentation accuracy, efficiency, and robustness, and is suitable for the segmentation and extraction of ski tracks in complex scenes, such as artificial snow fields.
Based on the improved algorithm of principal component analysis, we propose a ski resort point cloud registration method to overcome the low precision and lengthy registration process of ski resort point clouds. A Euclidean distance between the normal vectors of the snow field point cloud and an angle between the normal vectors are used to determine the feature points. The matching point set and corresponding pairing relationship are then calculated based on the feature, the point histogram, and the principal component analysis algorithm. The coarse registration is completed by combining the unit quaternion method. Also, the KD tree is utilized to aid in the iterative process of the ICP algorithm and to complete the fine registration of the point cloud data in the snowfield. Comparing the proposed new algorithm with NDT and ICP algorithms, the proposed new algorithm has significantly improved speed and accuracy in registering point clouds in snow fields.
Due to the high reflectivity of ski resort, the high hardness of artificial snow, the interference of light change, environmental noise and stony trees on the ground, and the problems of low registration efficiency and poor accuracy of the existing ICP algorithm, in this paper, a point cloud registration algorithm for artificial snow field based on symmetric point plane structure is proposed. Compared with the traditional point-to-point and point-to-face methods, we propose a symmetric point-to-plane structure, which can reduce the computational complexity and improve the computational efficiency and convergence speed. Based on the above theory, two groups of point cloud data of different ski resorts are selected for experiments. The experimental results show that the proposed method can be used to process the point cloud data of complex ski resorts in different environments and locations, in terms of registration speed and accuracy, good performance has been achieved.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.