基于VR全景图技术的乡村景观视觉评价偏好研究

发布时间：2018-04-11 03:24

本文选题：VR全景图 + 乡村景观　；参考：《北京林业大学学报》2016年12期

【摘要】：乡村景观评价是乡村景观规划的基础。为了更好地了解和研究乡村景观视觉环境的影响因素,本文首次将VR全景图技术应用于景观评价的研究中,进行了新技术应用于风景园林领域的实践尝试。利用VR全景图技术对空间、场景高度还原仿真的特点,改进传统的SBE法,以城市青年人群为例,通过观测实验并进行打分,从而得出原始数据,对乡村景观视觉评价及景观要素偏好进行研究。首先,使用改进后的SBE法与SD法构建了乡村景观视觉评价模型,并确定了4个显著影响指标(建筑肌理感、植物种类、道路形态和卫生状况)。其次,根据VR全景图场景中偏好视域截图,进而从截图中归纳提取出9个景观偏好指标。利用线性回归的方法,分别研究这9个景观偏好指标的偏好概率与该指标价值分数、所在场景SBE得分之间是否存在函数关系,并得出三者之间互相作用的模型。得出建筑特色指标与水环境指标的偏好概率与其自身的价值分数有显著关系;建筑肌理感、农田肌理感的偏好概率受其价值分数与场景SBE值共同影响;自然视野与其他视觉焦点的偏好概率受到场景SBE值影响更为显著;道路形态、道路材料和植物种类偏好概率与其自身价值分数、场景SBE值并无明显关系。最后,分析出城市青年人群会产生此种偏好的原因,并希望能够对未来乡村景观建设起到指导作用。
[Abstract]:Rural landscape evaluation is the basis of rural landscape planning.In order to better understand and study the factors affecting the visual environment of rural landscape, this paper first applied VR panorama technology to landscape evaluation, and made a practical attempt to apply new technology to landscape architecture.Using VR panoramic image technology to simulate the space and scene highly, the traditional SBE method is improved. Taking the young people in the city as an example, the original data are obtained by observing the experiment and scoring.The visual evaluation of rural landscape and the preference of landscape elements were studied.Firstly, the improved SBE method and SD method were used to construct a visual evaluation model of rural landscape, and four significant indexes (architectural texture, plant species, road shape and sanitary condition) were determined.Secondly, according to the view screenshots of VR panoramic scenes, nine landscape preference indexes are extracted from the screenshots.Using the linear regression method, the relationship between the preference probability and the value fraction of the nine landscape preference indexes and the SBE score of the scene is studied, and the model of interaction among the nine indexes is obtained.It is concluded that the preference probability of building characteristic index and water environment index has significant relationship with its own value score, and the preference probability of building texture feeling and farmland texture feeling is influenced by its value score and scene SBE value.The preference probability of natural visual field and other visual focus was significantly affected by the SBE value of scene, while the preference probability of road shape, road material and plant species had no significant relationship with their own value fraction and SBE value of scene.Finally, the reasons of this preference are analyzed, and it is hoped that it can guide the rural landscape construction in the future.
【作者单位】：北京林业大学园林学院;英属哥伦比亚大学统计学院;
【基金】：国家自然科学基金项目(31670704) 北京市共建项目(2016GJ03) 北京林业大学高水平全英文研究生核心课程项目(YWKC15016) 北京市园林绿化局计划项目(SYJYY2015002)
【分类号】：TU982.29

【相似文献】