XWXu Wang

Xu Wang

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

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Xu Wang has a total of 14 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    A novel method for ULS-TLS forest point cloud registration based on height context descriptor

    by Xusong Dai, Hanwen Qi, Xiaochen Wang, Yaozhan Xu, Qinghu Jiang, Qingjun Zhang, Xu Wang, Jianchang Chen, Guangzu Liu, Xinlian Liang


  • Open Access

    Individual Tree Segmentation From Airborne LiDAR Data Based on Automatic Treetop Detection and Simulated Stem-Branch Points Compensation

    by Qingjun Zhang, Jiale Chen, Hanwen Qi, Xu Wang, Xinlian Liang

    Accurate retrieval of tree structural attributes is essential for effective forest management and resource assessment. Airborne laser scanning (ALS) provides detailed threedimensional data from a top-down perspective, enabling the analysis of forest canopy structures at a large-scale. However, due to its limited penetration through dense canopies, ALS point clouds under the crown is often sparse or even missing, making individual tree segmentation particularly challenging. Therefore, we propose a novel individual tree segmentation method based on treetop detection and point compensation. First, a center shift algorithm (CSA) is proposed to detect treetop candidates, which are then refined to identify reliable treetops through geometricfeature analysis of these candidates' neighborhoods. Next, simulated stem-branch points are compensated with crown shape constraints to reconstruct sparse under-canopy structures. Finally, individual trees are segmented via a shortest-path algorithm applied to a graph that represents connectivity between adjacent points. The proposed method is evaluated using a public available dataset. The average extraction rate, completeness of treetop detection, and R2 for tree height estimation were 108%, 51%, and 0.93, respectively. In comparison to existing methods, the obtained the proposed method gave a similar F-Score as the best method, i.e., only 1% lower, but achieves a significantly better extraction rate, demonstrating superior overall performance. The proposed method enables accurate individual tree parameter extraction, advancing forest inventory efficiency and providing a foundation for ecological studies.


  • Automated Registration of Terrestrial Point Clouds Through Ground Overlapping Searching in Forests

    by Xusong Dai, Xinlian Liang, Hanwen Qi, Jianchang Chen, Xu Wang, Xiaochen Wang, Qingjun Zhang, Jian Zhang

    Terrestrial laser scanning (TLS) technology has been demonstrated to be able to measure forest structure fast, non-destructively, and with high precision. The registration of point cloud data from different scans is a prerequisite for an in-depth understanding of forest structure. Currently, automated registration methods for forest point clouds typically rely on tree attributes (such as tree position and stem diameter). However, these methods are often suitable for easy forest conditions, where the occlusion effects are less significant and the overlapped areas between scans are large. This paper proposes an automated point cloud registration method utilizing ground points without the need for extracting individual tree attributes. In the state-of-the-art hardware setup, the scanner automatically maintains a level position during data acquisition. Therefore, the differences in z-coordinates between the corresponding points in different scans should be equal. Leveraging this property and the Rhombic Region Correspondence Principle (RRCP), the proposed method identifies the overlapping rhombic region in the target and reference TLS data, which is located in the middle of the two scanning positions in the plot. Points within the overlapping rhombic region are used as registration primitives. Experiments were carried out in 24 plots with diverse stem densities, tree species and altitudes located in two test areas in Jiangxi and Zhejiang provinces, China. The resulting average pointwise error, average translation error, and average rotation error of TLS-TLS forest point cloud registrations are 5.77 cm, 4.92 cm, and 3.37 mrad. The results showed the potential of applying ground points for scan-to-scan registration.