Impact of Physical Features on Visual Walkability Perception in Urban Commercial Streets by Using Street-View Images and Deep Learning
Urban commercial streets are a crucial component of urban life, serving as the central hubs of commercial activity and providing vital spaces for both residents and visitors to engage in various activities. Walkability is commonly used as a key indicator of environmental quality, playing a significa...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-12-01
|
| Series: | Buildings |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2075-5309/15/1/113 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841549362620006400 |
|---|---|
| author | Gonghu Huang Yiqing Yu Mei Lyu Dong Sun Bart Dewancker Weijun Gao |
| author_facet | Gonghu Huang Yiqing Yu Mei Lyu Dong Sun Bart Dewancker Weijun Gao |
| author_sort | Gonghu Huang |
| collection | DOAJ |
| description | Urban commercial streets are a crucial component of urban life, serving as the central hubs of commercial activity and providing vital spaces for both residents and visitors to engage in various activities. Walkability is commonly used as a key indicator of environmental quality, playing a significant role in improving residents’ health, community interaction, and environmental quality of life. Therefore, promoting the development of a high-quality walking environment in commercial districts is crucial for fostering urban economic growth and the creation of livable cities. However, existing studies predominantly focus on the impact of the built environment on walkability at the urban scale, with limited attention given to commercial streets, particularly the influence of their physical features on walking-need perceptions. In this study, we utilized Google Street-View Panorama (GSVP) images of the Tenjin commercial district and applied the Semantic Differential (SD) method to assess four walking-need perceptions of visual walkability perception, including usefulness, comfort, safety, and attractiveness. Additionally, deep-learning-based semantic segmentation was employed to extract and calculate the physical features of the Tenjin commercial district. Correlation and regression analysis were used to investigate the impact of these physical features on the four walking-need perceptions. The results showed that the different walking-need perceptions in the Tenjin commercial district are attractiveness > safety > comfort > usefulness. Furthermore, the results show that there are significant spatial distribution differences in walking-need perceptions in the Tenjin commercial district. Safety perception is more prominent on primary roads, all four walking-need perceptions in the secondary roads at a high level, and the tertiary roads have generally lower scores for all walking-need perceptions. The regression analysis indicates that walkable space and the landmark visibility index have a significant impact on usefulness, street cleanliness emerges as the most influential factor affecting safety, greenness is identified as the primary determinant of comfort, while the landmark visibility index exerts the greatest influence on attractiveness. This study expands the existing perspectives on urban street walkability by focusing on street-level analysis and proposes strategies to enhance the visual walkability perception of commercial streets. These findings aim to better meet pedestrian needs and provide valuable insights for future urban planning efforts. |
| format | Article |
| id | doaj-art-92af06de2a3e4145b23c86512b626a19 |
| institution | Kabale University |
| issn | 2075-5309 |
| language | English |
| publishDate | 2024-12-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Buildings |
| spelling | doaj-art-92af06de2a3e4145b23c86512b626a192025-01-10T13:16:05ZengMDPI AGBuildings2075-53092024-12-0115111310.3390/buildings15010113Impact of Physical Features on Visual Walkability Perception in Urban Commercial Streets by Using Street-View Images and Deep LearningGonghu Huang0Yiqing Yu1Mei Lyu2Dong Sun3Bart Dewancker4Weijun Gao5Faculty of Environmental Engineering, The University of Kitakyushu, Kitakyushu 808-0135, JapanFaculty of Environmental Engineering, The University of Kitakyushu, Kitakyushu 808-0135, JapanSchool of Art and Design, Shenyang Jianzhu University, Shenyang 110168, ChinaSchool of Architecture and Urban Planning, Shenyang Jianzhu University, Shenyang 110168, ChinaFaculty of Environmental Engineering, The University of Kitakyushu, Kitakyushu 808-0135, JapanFaculty of Environmental Engineering, The University of Kitakyushu, Kitakyushu 808-0135, JapanUrban commercial streets are a crucial component of urban life, serving as the central hubs of commercial activity and providing vital spaces for both residents and visitors to engage in various activities. Walkability is commonly used as a key indicator of environmental quality, playing a significant role in improving residents’ health, community interaction, and environmental quality of life. Therefore, promoting the development of a high-quality walking environment in commercial districts is crucial for fostering urban economic growth and the creation of livable cities. However, existing studies predominantly focus on the impact of the built environment on walkability at the urban scale, with limited attention given to commercial streets, particularly the influence of their physical features on walking-need perceptions. In this study, we utilized Google Street-View Panorama (GSVP) images of the Tenjin commercial district and applied the Semantic Differential (SD) method to assess four walking-need perceptions of visual walkability perception, including usefulness, comfort, safety, and attractiveness. Additionally, deep-learning-based semantic segmentation was employed to extract and calculate the physical features of the Tenjin commercial district. Correlation and regression analysis were used to investigate the impact of these physical features on the four walking-need perceptions. The results showed that the different walking-need perceptions in the Tenjin commercial district are attractiveness > safety > comfort > usefulness. Furthermore, the results show that there are significant spatial distribution differences in walking-need perceptions in the Tenjin commercial district. Safety perception is more prominent on primary roads, all four walking-need perceptions in the secondary roads at a high level, and the tertiary roads have generally lower scores for all walking-need perceptions. The regression analysis indicates that walkable space and the landmark visibility index have a significant impact on usefulness, street cleanliness emerges as the most influential factor affecting safety, greenness is identified as the primary determinant of comfort, while the landmark visibility index exerts the greatest influence on attractiveness. This study expands the existing perspectives on urban street walkability by focusing on street-level analysis and proposes strategies to enhance the visual walkability perception of commercial streets. These findings aim to better meet pedestrian needs and provide valuable insights for future urban planning efforts.https://www.mdpi.com/2075-5309/15/1/113commercial streetsphysical featuresvisual walkability perceptionstreet-view imagesdeep learning |
| spellingShingle | Gonghu Huang Yiqing Yu Mei Lyu Dong Sun Bart Dewancker Weijun Gao Impact of Physical Features on Visual Walkability Perception in Urban Commercial Streets by Using Street-View Images and Deep Learning Buildings commercial streets physical features visual walkability perception street-view images deep learning |
| title | Impact of Physical Features on Visual Walkability Perception in Urban Commercial Streets by Using Street-View Images and Deep Learning |
| title_full | Impact of Physical Features on Visual Walkability Perception in Urban Commercial Streets by Using Street-View Images and Deep Learning |
| title_fullStr | Impact of Physical Features on Visual Walkability Perception in Urban Commercial Streets by Using Street-View Images and Deep Learning |
| title_full_unstemmed | Impact of Physical Features on Visual Walkability Perception in Urban Commercial Streets by Using Street-View Images and Deep Learning |
| title_short | Impact of Physical Features on Visual Walkability Perception in Urban Commercial Streets by Using Street-View Images and Deep Learning |
| title_sort | impact of physical features on visual walkability perception in urban commercial streets by using street view images and deep learning |
| topic | commercial streets physical features visual walkability perception street-view images deep learning |
| url | https://www.mdpi.com/2075-5309/15/1/113 |
| work_keys_str_mv | AT gonghuhuang impactofphysicalfeaturesonvisualwalkabilityperceptioninurbancommercialstreetsbyusingstreetviewimagesanddeeplearning AT yiqingyu impactofphysicalfeaturesonvisualwalkabilityperceptioninurbancommercialstreetsbyusingstreetviewimagesanddeeplearning AT meilyu impactofphysicalfeaturesonvisualwalkabilityperceptioninurbancommercialstreetsbyusingstreetviewimagesanddeeplearning AT dongsun impactofphysicalfeaturesonvisualwalkabilityperceptioninurbancommercialstreetsbyusingstreetviewimagesanddeeplearning AT bartdewancker impactofphysicalfeaturesonvisualwalkabilityperceptioninurbancommercialstreetsbyusingstreetviewimagesanddeeplearning AT weijungao impactofphysicalfeaturesonvisualwalkabilityperceptioninurbancommercialstreetsbyusingstreetviewimagesanddeeplearning |