摘要
现有的景观模型特征点重标定方法易受到操作者主观的影响,存在重标定准确率较低的问题。为解决上述问题,研究设计了新的生态景观三维模型关键特征点重标定方法。以生态景观三维模型为基础搭建三维模型尺度空间,并提取三维模型关键特征点。通过插值、剔除低对比度以及降低边缘效应的方式过滤关键特征点,根据差分值计算关键特征点梯度大小与方向,并制定关键特征点描述符。采用手工方式初次标定生态景观三维模型的关键特征点。由于初标定准确率较低,故采用高斯曲率对关键特征点进行重标定。实验结果显示,与现有重标定方法平均数值相比较,新方法的高斯曲率绝对值上升、外接圆直径减少,表明应用新方法后,对关键特征点重标定的准确率更高。
The subjective will of the operator affects the re-calibration of landscape model feature points,resulting in low re-calibration accuracy.Therefore,this paper designs the re-calibration method of the key feature points of the three-dimensional model of the ecological landscape.Based on the three-dimensional model of the ecological landscape,the scale space of the three-dimensional model was established,and the key feature points of the three-dimensional model were also extracted.The key feature points were filtered via interpolation,eliminating low contrast and reducing the edge effect.Based on the difference value,the gradient size and direction of the key feature points were calculated,and the descriptor of the key feature points was formulated.The manual method was used to calibrate the key feature points of the ecological landscape 3 D model for the first time.Gaussian curvature was adopted to re-calibrate key feature points due to low initial calibration accuracy.Experimental results show that the method can improve the absolute value of Gaussian curvature,reduce the circumscribed circle diameter and improve the re-calibration accuracy of key feature points.
作者
李珂
张晓波
LI Ke;ZHANG Xiao-bo(Shenzhen University,Shenzhen Guangdong 518000,China;Shandong Jianzhu University,Jinan Shandong 250101,China)
出处
《计算机仿真》
北大核心
2022年第11期366-370,共5页
Computer Simulation
基金
平面设计艺术在教学中的实践和应用研究(JKS83437)。
关键词
生态景观模型
关键特征点
特征重标定
差分值计算
高斯曲率
Ecological landscape model
Key feature points
Feature re-calibration
Differential calculation
Gaussian curvature
