| As an important blue-green space in cities,waterfront greenways undertake multiple functions such as leisure and recreation,landscape appreciation,cultural and scientific popularization,and physical and mental healing for urban residents.However,the current flow of people and recreational functions of different sections of the greenway are quite different,with many tourists,good reachability and good hydrophilicity.The locations that are easy to become recreational hotspots often do not bring high-quality visual landscape effects,which greatly affects the overall function of the greenway.At the same time,there is a lack of prior screening and evaluation of recreational hotspots in current research and practical projects,and it is unclear where recreational hotspots are located,nor has targeted visual landscape optimization and improvement been carried out based on the results of recreational hotspots.In this paper,the research content is clarified by reading the literature related to greenway recreation hotspots,visual landscape and landscape architecture big data.According to the upper planning,completion and opening conditions,the research objects are initially selected,which are divided into urban lakeside type,rural lakeside type and riverside type.According to the principle of relatively balanced spatial distribution,8 waterfront greenways(groups)are finally selected as the research objects.Based on the field survey of 8 waterfront greenways,panoramic photos and location coordinates were obtained.Using Baidu thermal map,POI,urban public transport,urban road network and other big data,nine indicators including actual visits,relative heat difference,POI density,POI diversity,the nearest bus station,reachability of vehicle traffic,hydrophilicity,space carrying capacity and node nature were determined from the two dimensions of actual characterization and potential characterization,Obtain a calculation model for greenway recreational hotspots using the entropy weight method,and analyze the distribution pattern of recreational hotspots.Using the combination of image semantic segmentation technology(FCN)and manual interpretation,the panoramic photos of selected recreation hotspots were analyzed from eight visual landscape factors,including green visual rate,sky openness,blue space ratio,community configuration ratio,debris interference,sky features,plant seasons,aquatic plant,and so on,to reveal the visual landscape effect of each recreation hotspot.The main conclusions are as follows:(1)Screen greenway recreational hotspots using 9 indicators from two dimensions:actual representation and potential representation.Use the entropy weight method to obtain the weights of each indicator.The results show that: node performance>hydrophilicity>POI density>relative heat difference>actual visits>POI diversity>space carrying capacity>vehicle traffic reachability>nearest bus stop,and the first four indicators account for 85.12% of the weight;(2)The overall score of recreational hotspots on the greenway indicates that the average scores of the Happiness Street greenway,Hanyang Jiangtan greenway,and Shahu greenway are in the top three,indicating that when all greenway research sites are compared simultaneously,recreational hotspots will appear at most on the three greenways mentioned above;The average value of the three greenways along the riverbank is 25.2,which is much higher than 13.6 for the urban lakeside type and 9.1for the suburban lakeside type;(3)Analyze the spatial distribution pattern of recreational hotspots in each greenway,which can be divided into "isolated POInt distribution","small aggregation distribution",and "linear distribution".The greenways in Yuehu,Mohe Lake,Hankou River Beach,Xingfu Street,Jinyin Lake,and Houguan Lake have the most isolated POInt distribution,while the small aggregation distribution in Shahu Lake has the most.The Hanyang Jiangtan Greenway is only distributed in a linear pattern,while the Happy Street Greenway is only distributed in isolated POInts;(4)Using image semantic segmentation technology and manual interpretation to determine 8 visual landscape factors for greenway visual landscape analysis,the results showed that in terms of green vision rate,the proportion of recreational hotspots in Houguanhu greenway with the best green vision rate is the highest,exceeding 20%;In terms of sky opening,the proportion of sky opening in the optimal range for recreational hotspots such as Shahu Greenway,Hankou Binjiang Greenway,Jinyin Lake Greenway,and Houguan Lake Greenway exceeds 50%;In terms of blue space ratio,the blue space of riverside greenways is significantly higher than that of suburban and urban lakeside greenways;In terms of community configuration ratio,the greenways that reach the optimal range of this factor are only Shahu Lake,Xingfu Street,Jinyin Lake,and Ink Lake;In terms of debris interference,Shahu Lake,Yuehu Lake,and Jinyin Lake are heavily disturbed by guardrails;In terms of astronomical characteristics,the Jinyin Lake Greenway has the largest number of astronomical recreational hotspots;In terms of plant seasonal landscape,only the recreational hotspots of 5 greenways have plant seasonal landscape,and the color leaf tree species are single;In terms of aquatic plant,the proportion of the urban lakeside type three greenway recreation hotspots with aquatic plant landscape effect ranks in the top three;(5)Propose optimization strategies and suggestions for the visual landscape of waterfront greenways from the spatial distribution of recreational hotspots at the macro planning level and the visual landscape factors of each greenway at the micro level.This study screened the recreational hotspots of waterfront greenways through multi-source data,and utilized image semantic segmentation technology combined with manual recognition to analyze the visual landscape in a targeted manner.The aim is to provide assistance for improving the visual landscape services of the recreational hotspots of waterfront greenways in Wuhan and constructing a more diverse and highquality waterfront greenway system. |
PDF Full Text Request