| Road construction is the foundation and important condition of economic development.Nowadays,China’s road construction is advancing by leaps and bounds,and the construction of road greening landscape has attracted more and more attention.The road greening landscape reflects the appearance of the city and is related to the daily life of citizens,Under the urban construction concept of "people’s cities are built by the people,and people’s cities are for the people",the demand for optimizing and transforming road greening landscape is increasing.Based on the insufficient time span and subjective quantification methods in the current research progress,the experiment investigated the regional road network of Pujiang town.From the perspective of pedestrians,30 sample points were set up in 11 urban secondary roads of Pujin community and Pujiang town urban section.The grey statistical method is used to screen landscape factors to reduce the subjectivity of landscape factor selection.The landscape factor data is objectively quantified,in which the deep learning full convolution network model(FCN)is used for semantic image segmentation.After that,The landscape visual evaluation of four seasons was carried out by using the beauty degree evaluation method(SBE),and the changes and characteristics of the beauty degree of the sample points were analyzed.Finally,Multiple regression analysis of SPSS 22.0 was carried out on the beauty evaluation results and landscape factor data,and the beauty evaluation model was established.Explain the influencing factors of road greening landscape in theory,and put forward scientific suggestions for the optimization of road greening landscape.The main results are as follows:(1)Through the beauty evaluation,the average value of greening landscape of secondary trunk road in Pujiang town is-0.04,and the overall beauty value is low.Among them,the negative value samples account for 60% of the 30 sample points,and the overall landscape effect needs to be improved.The values of beauty degree are 0.03 in autumn,0.02 in summer,0.00 in spring and-0.21 in winter.Winter and spring have a great impact on the beauty degree of the four seasons.Among them,the samples with large fluctuations in seasonal differences account for 40% of the overall sample points.There is room for improvement for the seasonal beauty difference of the landscape.(2)The landscape factors that have the most significant impact on the beauty of road greening in different seasons are obtained.In the autumn road landscape,the green viewing rate,the three-tier configuration level,the two-tier configuration level of the central reservation have the most significant and positive correlation on the beauty value.In the winter road landscape,the building proportion,the two-tier configuration level,the three-tier configuration level of the central reservation have the most significant impact on the beauty degree,and the building proportion is negatively correlated with the beauty degree.In spring,sidewalk planting with two-layer configuration,wide ovate roadside trees,umbrella roadside trees,color leaf and flowering plants have the most significant impact on the beauty degree,among which the umbrella roadside trees are negatively correlated with the beauty degree.In summer,the green visual rate and plant species have the most significant impact on the beauty value.(3)Different gender,professional occupation and long-term living people in specific areas have significant correlation in the results of road landscape greening evaluation,and the aesthetic difference of personnel at these three levels is small.Among them,the evaluation results of the same professionals have the highest correlation,indicating that professional learning will have a slight impact on aesthetics.In the future,the selection of researchers in landscape evaluation research should avoid being limited to the selection of their own major or similar professional groups as far as possible.(4)This paper analyzes the typical problems of roads in the experimental area,and puts forward targeted solutions to the improvement of road landscape greening.The central reservation model,which emphasizes the ornamental ability of shrubs in four seasons,the permeability of trees in winter and the ornamental ability of herbs in early spring,is recommended;Emphasis should be placed on increasing the pattern of vehicle greenbelt of white,purple and red flowering shrubs and herbs;It emphasizes the greenbelt pattern of sidewalk planting based on the color change of tree branches and leaves and the ornamental flowers and fruits of shrubs;It emphasizes the roadside greenbelt model with the theme of evergreen tree enclosure,lawn space shaping,colored leaf and flowering plants.Finally,combined with the spatial planning and current situation of innovative industrial community,country park area and residential community in the study area,the improvement practice of road greening landscape is carried out. |
PDF Full Text Request