三维景观规划图像特征增强方法仿真研究  

Simulation Research on Image Feature Enhancement Method of 3D Landscape Planning

作　　者：赵立恒[1] 

机构地区：[1]北华大学师范分院,吉林吉林132013

出　　处：《计算机仿真》2017年第12期407-410,共4页Computer Simulation

摘　　要：对三维景观的高效规划,能够有效提升景观规划设计的视觉效果。对景观规划中三维图像特征的增强,需要依据获取图像比例信息特征增强阈值,给出每个像素的邻域信息,完成对图像特征增强。传统方法找到景观图像相应局部特征点的仿射变换参数,使目标函数达到最小,但忽略了给出像素的邻域信息,导致特征增强效果不理想。提出基于正视图的景观三维图像特征增强方法。利用图像的直方图统计信息原理从不同的方位分别对景观三维图像进行投影,对图像进行二维傅立叶变换,并进行低频采样从而得到图像的特征向量,获取图像比例信息特征增强阈值,并给出图像中的每个像素的邻域信息,完成对景观三维图像特征增强。仿真证明,所提方法景观三维图像特征增强的幅度较大,为提升对景观三维图像的质量奠定基础。This paper proposes a feature enhancement method for 3 D sight image based on elevation view. First of all, the method projects 3D sight image from different directions by using statistical information principle of histogram of image, then carries out 2D Fourier transformation for image and obtains feature vector of image by low frequency sampling, and then obtains feature enhancement threshold of image proportional information and gives neighborhood information of each pixel in the image. At last, feature enhancement of 3D sight image is achieved. The simulation results demonstrate that the method in this paper has a large range of feature enhancement for 3D sight image feature. It can improve quality of 3D sight image.

关 键 词：景观规划 三维图像 特征增强 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]