基于视觉感知的传统村落街巷风貌研究

Research on the Street and Alley Style of Traditional Villages Based on Visual Perception

摘要:

街巷风貌是传统村落景观的重要组成部分,为探究传统村落街巷风貌的游客视觉感知特征,采用眼动实验法、深度学习图像语义分割和主观问卷等方法,对街巷空间风貌进行定性定量评价。结果表明:游客视线多集中于建筑、景观节点和道路尽头,游客对于标志物和景观节点平均注视时长更长;舒适度与秩序感、植物配置、色彩、协调性、视觉层次均呈现显著的正相关;平均注视次数与绿视率、植物配置和色彩呈现显著的正相关,与建筑围合度呈现显著的负相关。游客对于不同要素比例的街巷场景的视觉感知存在显著差异。

Abstract:

The street and alley appearance is an important component of the traditional village landscape. To explore the visual perception characteristics of tourists towards the street and alley appearance in traditional villages, this paper adopts eye movement experiments, deep learning image semantic segmentation and subjective questionnaires to conduct qualitative and quantitative evaluations of the street and alley space appearance. The results show that the tourists’ gazes are mostly concentrated on buildings, landscape nodes and the end of the road, and tourists’ average fixation time on landmarks and landscape nodes is longer. Comfort and orderliness, plant configuration, color, coordination and visual hie-rarchy all show significant positive correlations. The average fixation count shows significant positive correlations with green view ratio, plant configuration and color, and a significant negative correlation with building enclosure degree. There are significant differences in tourists’ visual perception of street and alley scenes with different element proportions.