Three-dimensional system modeling and design of ecological garden landscape based on the interlaced spatial pattern of light and shadow

Article type: Research Article

Authors: Chen, Chen

Affiliations: School of Environmental Arts and Design, Wuxi Vocational Institute of Arts and Technology, Wuxi, Jiangsu 214206, China | E-mail: chenc-2t@163.com

Correspondence: [*] Corresponding author: School of Environmental Arts and Design, Wuxi Vocational Institute of Arts and Technology, Wuxi, Jiangsu 214206, China. E-mail: chenc-2t@163.com.

Abstract: The layout and design of ecological landscaping is an important part of the construction and development of modern cities. In the 3D reconstruction of the spatial pattern of the light and shadow interlaced zone of the ecological landscape, the complexity and particularity of the ecological landscape structure make it difficult for the three-dimensional reconstruction stereo matching set to meet the accuracy requirements, and the quality 3D image construction cannot meet the requirements of landscape planning. Based on the principle of binocular stereo vision, a regional feature stereo matching algorithm (rsurf) is used to improve the accuracy of feature matching. Considering that the algorithm is easy to filter out the detailed features of the image, the improved RANSAC algorithm is used to filter the matching results. The experimental results show that in the matching cost test of the optimal matching window, the 15 × window neighborhood has the lowest matching cost, and the generated value in the 100 × 100 source window is 0.824. In the test after matching and fusion, the rsurf algorithm is superior to the surf algorithm in both RMS and PMS error performance, and can better meet the requirements of 3D reconstruction of the binocular vision system. The research content has an important reference for the application of landscape visualization 3D technology, and improves the overall layout effect of landscape landscape.

Keywords: Ecological garden landscape, criss-cross area, spatial pattern, improved RANSAC algorithm, stereo matching

DOI: 10.3233/JCM-226712

Journal: Journal of Computational Methods in Sciences and Engineering, vol. 23, no. 3, pp. 1503-1516, 2023