基于视觉感知的乡村居住景观视觉敏感度特征提取  

Feature extraction of visual sensitivity of rural residential landscape based on visual perception

作　　者：邹初红 ZOU Chu-hong(Academy of Art Design,West Anhui University,Anhui Lu'an 237012,China)

机构地区：[1]皖西学院艺术学院,安徽六安237012

出　　处：《齐齐哈尔大学学报（自然科学版）》2021年第4期66-70,84,共6页Journal of Qiqihar University(Natural Science Edition)

基　　金：皖西地区乡村人居环境研究项目(SK2019A0442)。

摘　　要：由于当前景观特征敏感度提取方法存在特征点定位误差较大的问题,提出一种基于视觉感知的乡村居住景观视觉敏感度特征提取方法。采集乡村居住景观视觉敏感度的遥感数据与矢量数据,其中遥感数据收集使用的遥感信息源为资源三号影像,收集的矢量数据则包括乡村居住景观的数字地形图等。使用ArcGIS 2.0与ENVI 3.4软件,对相关数据实施预处理。预处理的环节包括图像融合、辐射校正、配准等环节。基于处理后的数据构建乡村居住景观区域视觉敏感度数据的数字高程模型。基于视觉感知对乡村居住景观视觉敏感度数字高程模型中的特征进行分析,通过SIFT算法分别对几种特征进行提取。通过进行对比实验,证明设计方法的特征点定位误差较小,应用价值很高。Given the current landscape characterized by the presence of the sensitivity of feature points extracted large error method,and therefore proposed rural residential landscape visual sensitivity characteristics based on visual perception extraction methods.Rural Landscape visual sensitivity acquisition live remote sensing data and vector data,wherein the remote sensing data collected by remote sensing information for the resource used in the 3 rd image vector data was collected includes a digital topographic map,etc.rural residential landscape.Use ArcGIS 2.0 and ENVI 3.4 software to preprocess related data.The preprocessing steps include image fusion,radiometric correction,and registration.The processed data to build a digital elevation model of rural residential area landscape visual sensitivity data.Analysis of rural residential landscape visual sensitivity digital elevation model based on visual perception characteristics,several features are extracted by SIFT algorithm respectively.By comparison experiments show features,in the design method of the error was small,a high application value.

关 键 词：视觉感知 乡村居住景观 图像融合 

分 类 号：P901[天文地球—自然地理学] TP391.41[自动化与计算机技术—计算机应用技术]