CN115271549A - A method of constructing a comprehensive evaluation index system of urban park space quality - Google Patents

Abstract

The invention discloses a method for constructing a comprehensive evaluation index system of urban park space quality, which belongs to the technical field of municipal construction and comprises the following steps; constructing an urban park material environment space quality evaluation index system; establishing an urban park perception environment space quality evaluation index system; determining the weight of indexes in each dimension and each scale by adopting an analytic hierarchy process, and finally constructing a comprehensive evaluation index system of the urban park space quality; the invention takes the integrated park as the key ecological background in the process of building the park city, adopts new data formed by massive street view data and other development data sources and new technical environment formed by deep learning in the field of artificial intelligence and the like to build a park space quality comprehensive evaluation index system with two dimensions of park material environment and perception environment and two dimensions of external community and internal community under each dimension, and realizes systematic evaluation combining quantification and qualification.

Description

A construction method of comprehensive evaluation index system for urban park space quality

technical field

The invention relates to the technical field of municipal construction, in particular to a method for constructing a comprehensive evaluation index system for urban park space quality.

Background technique

Parks are an important part of urban landscape, and the evaluation of park space quality is of great value to urban planning and management. At present, the evaluation methods of park space quality mainly include classic methods based on questionnaire survey and regression analysis, and new technology methods based on "remote desktop evaluation tool for public open space quality". These methods are limited by time and manpower, and it is difficult to achieve detailed and full coverage of the entire city through questionnaire surveys and field surveys. Large-scale and refined evaluation of park space quality has become an urgent problem to be solved.

Contents of the invention

The purpose of the present invention is to provide a method for constructing a comprehensive evaluation index system of urban park space quality, so as to solve the problems raised in the above-mentioned background technology.

To achieve the above object, the present invention provides the following technical solutions:

A method for constructing a comprehensive evaluation index system for urban park space quality, comprising the following steps;

The spatial quality of urban parks is divided into two dimensions of physical environment and perceived environment, and two scales of external community and internal community;

Based on the above-mentioned two dimensions and two scales, construct an evaluation index system for the quality of urban park physical environment space;

Based on the above-mentioned two dimensions and two scales, an evaluation index system for the perceived environmental space quality of urban parks is constructed;

The analytic hierarchy process is used to determine the weight of indicators in each dimension and scale, and finally build a comprehensive evaluation index system for urban park space quality.

As a further technical solution of the present invention: the construction of the urban park physical environment space quality evaluation index system specifically includes: the construction of external community scale indicators under the park physical environment and the construction of internal scale indicators under the park physical environment.

As a further technical solution of the present invention: the construction of external community-scale indicators under the physical environment of the park includes: selecting external community-scale evaluation indicators based on the literature review method and Delphi method, and the external community-scale evaluation indicators include: standardized angle integration density, bus station density, distance from the subway entrance, functional density, functional mix, land use properties of the surrounding plots, and development intensity of the surrounding plots, the selected indicators are quantified by using space syntax and GIS spatial analysis and statistical methods, and the park material External community-scale indicators in context.

As a further technical solution of the present invention: the construction of the internal scale index under the physical environment of the park includes: selecting the internal scale evaluation index based on the literature review method and the Delphi method, and the internal scale evaluation index includes: standardized angle integration degree, matching The level of service facilities, functional diversity and water cleanliness are quantified by using space syntax and on-site investigation methods to obtain internal scale indicators in the park's physical environment.

As a further technical solution of the present invention: the construction of the urban park perception environment space quality evaluation index system includes: the construction of the external community scale index in the park perception environment and the construction of the internal scale index in the park perception environment.

As a further technical solution of the present invention: the construction of external community-scale indicators under the park perception environment includes: selecting external community-scale evaluation indicators based on the literature review method and Delphi method, and the external community-scale evaluation indicators include green viewing rate, Sky opening degree, interface enclosure degree, motorization degree, sidewalk perception degree, building perception degree and the proportion of human flow, through the Google Deeplab convolutional neural network technology based on deep learning in the field of artificial intelligence and the Cityscapes evaluation data set for indicators Quantification is carried out to obtain the external community scale indicators in the park's perceived environment.

As a further technical solution of the present invention: the construction of the internal scale index under the park perception environment includes: subdividing the internal space of the park into four types: garden road space, waterfront space, square space and architectural sketch space, more detailed An evaluation index system for the perceived environmental space quality of parks is constructed in an efficient way. Based on the literature review method, Delphi method, landscape beauty evaluation method and statistical analysis method, internal scale evaluation indicators are selected, including: visual color aesthetics, ground pavement aesthetics , Plant species or richness of landscape elements, green viewing rate, sky openness, perception of roads and squares, perception of buildings and structures, and proportion of people flow, quantify the indicators through questionnaire survey, Deeplab and Cityscapes technology, and get the perceived environment of the park Next, the internal scale indicators. Although the aesthetics evaluation index is subjective, it collects street view photos of sampling points and incorporates them into the evaluation index system in the form of scoring to ensure its integrity.

As a further technical solution of the present invention: the weight value of each index in the evaluation index system is obtained by using the AHP and the Delphi method, specifically using YAAHP software to construct the levels of each index in the evaluation index system, and then through two The importance of the two comparative indicators is given the corresponding value, and after the consistency test is repeated 20 times, the average weight value of each indicator in the evaluation index system is obtained.

Compared with the prior art, the beneficial effect of the present invention is: the present invention takes the comprehensive park as the key ecological background in the park city construction process, adopts massive street view data and other development data sources to form new data and deep learning in the field of artificial intelligence Based on the new technology environment composed of the park, the two dimensions of the park's physical environment and the perceived environment, as well as the comprehensive evaluation index system of the park's spatial quality at the external community and internal scales under each dimension, realized the systematic evaluation combining quantitative and qualitative.

Description of drawings

Figure 1 is a schematic diagram of the evaluation index system of the park's physical environment space quality.

Figure 2 is a schematic diagram of the evaluation index system of the perceived environmental space quality of the park.

Figure 3 is a schematic diagram of the comprehensive evaluation index system of urban park space quality.

Detailed ways

The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the protection scope of the present invention.

As shown in Figure 1-3, a method for constructing a comprehensive evaluation index system for urban park space quality includes the following steps;

Firstly, the spatial quality of urban parks is divided into two dimensions of physical environment and perceived environment, and two scales of external community and internal community;

Secondly, based on the above-mentioned two dimensions and two scales, construct an evaluation index system for the quality of urban park physical environment space;

Thirdly, based on the above-mentioned two dimensions and two scales, construct an evaluation index system for the perceived environmental space quality of urban parks;

Finally, the analytic hierarchy process is used to determine the weights of indicators in each dimension and scale, and finally a comprehensive evaluation index system for urban park space quality is constructed.

Embodiment 2, on the basis of embodiment 1,

The construction of an evaluation index system for the physical environment and space quality of urban parks includes the following steps:

1) Contents of evaluation indicators at the community scale outside the park:

①Normalized angle integration degree (NAIN): import the preprocessed urban road network vector map into Depthmap for line segment model analysis, and characterize the accessibility of the studied park by calculating the global NAIN.

②Bus station density: Through GIS spatial analysis technology, the center of the research park land is used as a circle point, and the radius of 1000 meters is used as a buffer zone, and the proportion of bus stations within the buffer zone is calculated to obtain the bus station density.

③The shortest walking distance between the entrances and exits of the park and the subway station entrance: Check the shortest walking distance between each entrance and exit of the park and the subway station entrance through Baidu network map viewing and distance measurement or GIS spatial analysis technology.

④Functional density: Through GIS spatial analysis technology, the center of the research park land is used as a circle point, and the radius of 1000 meters is used as a buffer zone, and the proportion of POI points within the buffer zone is calculated to obtain the POI point density.

⑤Functional mixing degree: Through GIS spatial analysis technology, the center of the research park land is used as a dot, and the radius of 1000 meters is used as a buffer zone, and the information entropy of various POI proportions within the buffer zone is calculated respectively to obtain the POI point mixing degree.

⑥Land use properties of the surrounding plots: take the center of the research park as a dot and use a radius of 1,000 meters as a buffer zone to determine the current land use properties of each plot within this range. 60%, then assign the land use property attribute of the plot to the researched park, and obtain the land use properties of the surrounding plots of the park.

⑦Development intensity of surrounding plots: Calculate the building area corresponding to the building surface data through ArcGIS software, and calculate the land area based on the 1000m buffer zone of the park. Intensity of plot development.

2) Contents of community-scale evaluation indicators within the park:

①Normalized angle integration degree (NAIN) and normalized angle integration degree (NACH): By using Depthmap to analyze the line segment model of the internal park roads under study, the global NAIN and NACH are obtained.

②Supporting service facility level: Through on-site surveys, the relevant indicators of supporting service facilities were collected: various parking lots, public toilets, commercial kiosks (including self-service shopping machines), seats, trash cans, teahouses, explanation services (including voice explanation system), guide signboards, alarm and help facilities (including self-service alarm facilities), automatic external defibrillator, use the dichotomy method to score the facilities according to the availability, quantity and quality of the facilities.

③Functional diversity: Through on-site surveys, the related indicators of functional diversity were collected: entertainment facilities for children, activity facilities for young and middle-aged people, and activity facilities for the elderly, and they were measured using a dichotomous method in terms of their function, quantity, and quality. table rating.

④Water body cleanliness: Through on-site survey, use the dichotomous method to score the scale from the presence or absence of water features, quantity and quality.

Embodiment 3. On the basis of Embodiment 2, establishing an evaluation index system for urban park perception environment space quality specifically includes the following steps:

1) The content of evaluation indicators at the community scale outside the park.

①Green viewing rate: take the percentage of vegetation as the green viewing rate. Calculated from street view photo data, the calculation formula is as follows:

L=P_vegetation/P In the formula, L is the green viewing rate, P_vegetationis the total pixel value of green vegetation in each photo, and P is the total pixel value of each photo.

②Sky openness: The sky openness is measured by the sky area ratio. Calculated from street view photo data, the calculation formula is as follows:

T= _Psky /P where T is the openness of the sky, Psky is the total pixel value of the sky in each photo, and P is the total pixel value of each photo.

③Interface enclosure: The total ratio of buildings, fences, walls, poles, signs and vegetation is taken as the interface enclosure. Calculated from street view photo data, the calculation formula is as follows:

J _interface = P _interface / P where J _interface is the interface enclosure, P _interface is the total pixel value of buildings, fences, walls, poles, signs and vegetation in each photo, and P is the total pixel value of each photo value.

④ Degree of motorization: The degree of motorization is measured by the sum of the proportion of roads and vehicles. Calculated from street view photo data, the calculation formula is as follows:

_{Jmotorization} =Pmotorization/P where _{Jmotorization} is the degree of _motorization , P _interface is the total pixel value of vehicles and vehicles in each photo, and P is the total pixel value of each photo.

⑤Sidewalk perception: take the proportion of sidewalk as the sidewalk perception. Calculated from street view photo data, the calculation formula is as follows:

R _sidewalk = P _sidewalk / P where R _sidewalk is the sidewalk perception, P _sidewalk is the total pixel value of the sidewalk in each photo, and P is the total pixel value of each photo.

⑥Building Perception: measure building perception by building ratio. Calculated from street view photo data, the calculation formula is as follows:

B _building = P _building / P where B _building is the perception of the building, P _sidewalk is the total pixel value of the building in each photo, and P is the total pixel value of each photo.

⑦Proportion of people flow: the number of people is used as the proportion of people flow. Calculated from street view photo data, the calculation formula is as follows:

R _people = P _people / P where R is the flow of _people , P is the total pixel value of the flow of _people in each photo, and P is the total pixel value of each photo.

2) Contents of the internal scale evaluation indicators of the park:

①Based on the walkable range, take an average of 10 meters as a sampling point, use the GPS tool to obtain the latitude and longitude coordinates of the point, and then take 180° street view photos from a human perspective, and take two photos for each sampling point, and get each 360° panoramic photos of sampling points.

②According to the characteristics of the built environment of the park, the internal space of the park is divided into four different types of spaces: garden road space, waterfront space, square space and architectural sketch space.

The garden road space refers to the various roads in the built environment of the park and the green space in the surrounding environment, and also includes landscape bridges and elevated plank roads with walking functions.

Waterfront space refers to the waterscape with a certain scale for people to visit in the built environment of the park and the green space in its surrounding environment, including the garden roads adjacent to the waterscape.

Square space refers to various landscape theme squares in the built environment of the park, distribution squares at the entrance and exit of the park, activity venues for children and the elderly, sports and fitness venues, open-air performance venues and other paved venues with a certain scale for people to use, including other Green spaces in the surrounding environment.

Architectural sketch space refers to the green space in the built environment of the park with certain functions and small architectural facilities or landscape facilities such as pavilions, corridors, frames, and pavilions, as well as the surrounding environment, and also includes the garden roads adjacent to the architectural sketches.

③According to the collection of large samples of street view photos of 4 different types of spaces in the parks in the central urban area, the quality of various types of spaces perceived by human judges and 6 evaluation indicators that may affect the quality were realized, namely the green viewing rate, the sky Openness, perception of roads and squares, perception of buildings and structures, interface enclosure, proportion of pedestrian flow.

④Based on the results of the above analysis, select the common evaluation indicators applicable to four different types of spaces, namely, the aesthetics of landscape color, the aesthetics of ground pavement, the richness of plant species or landscape elements, and manually evaluate the street view photos of the parks at each sampling point. Statistical correlation analysis and regression analysis were performed on the perceived quality of park space and the three evaluation indicators that may affect the quality.

⑤Based on the evaluation indicators obtained in steps ③ and ④, which are applicable to various types of spaces and the common evaluation indicators applicable to 4 different types of spaces, finally form 4 different types of spaces (garden road space, Waterfront space, square space and architectural sketch space) evaluation index content.

Garden road space quality evaluation index content:

Green viewing rate: take the proportion of vegetation as the green viewing rate. Calculated from street view photo data, the calculation formula is as follows:

L=P _vegetation /P where L is the green viewing rate, P _vegetation is the total pixel value of green vegetation in each photo, and P is the total pixel value of each photo.

Sky Openness: The sky openness is measured by the sky area ratio. Calculated from street view photo data, the calculation formula is as follows:

T= _Psky /P where T is the openness of the sky, Psky is the total pixel value of the sky in each photo, and P is the total pixel value of each photo.

Perception of road squares: take the proportion of road squares as the perception of road squares. Calculated from street view photo data, the calculation formula is as follows:

R=P _{road square} /P where R is the road square perception, P _{road square} is the total pixel value of the road square in each photo, and P is the total pixel value of each photo.

Building perception: Building perception is measured by building ratio. Calculated from street view photo data, the calculation formula is as follows:

B _building = P _building / P where B _building is the perception of the building, P _sidewalk is the total pixel value of the building in each photo, and P is the total pixel value of each photo.

Interface enclosure: The total proportion of buildings, fences, walls, poles, signs and vegetation is taken as the interface enclosure. Calculated from street view photo data, the calculation formula is as follows:

J _interface = P _interface /P formula where J _interface is the interface enclosure, P _interface is the total pixel value of buildings, fences, walls, poles, signs and vegetation in each photo, and P is the total pixel value of each photo value.

Landscape color aesthetics: It refers to the psychological changes caused by the stimulation of human visual system by the landscape color, and the color factor is considered from the change of the four seasons of the landscape.

Through the street view data, using the deep learning technology in the field of artificial intelligence, the raters with professional background first scored 1 to 3 points on the street view data of some parks according to four standards (the specific standards are attached), 1 is low and 2 points Average, 3 points are high, and the park street view data after scoring are used as sample data. In TensorFlow, the second-generation open source artificial intelligence learning system developed by Google, the above park street view data and corresponding score values are called in batches using the Python programming language , through deep learning technology to obtain its nonlinear complex feature selection, so as to have the ability of automatic batch scoring for most of the remaining park street view data, that is, through deep learning model training on part of the park street view data, it is realized that most of the remaining park street view data Data autonomous batch scoring.

evaluation standard:

(1) Scoring criteria for the color aesthetics of garden road space landscape: 0.5 points for garden roads with a cool color system and a single color, 1 point for a garden road with a neutral color system and a single color, 1.5 points for other garden roads; single green color 0.5 points for green color equal to two, 1 point for other greening conditions, 1.5 points.

(2) Scoring standard for waterfront space landscape color aesthetics: 0.5 points for water body that is black, 1 point for turbid water body that is gray, 1.5 points for clear water body that is colorless and transparent; 0.5 point for single green color, equal to two green colors 1 point for greening and 1.5 points for other cases of greening.

(3) Scoring criteria for the color aesthetics of the square space landscape: 0.5 points for a single square color, 1 point for two square colors, and 1.5 points for the rest of the square; 0.5 points for a single green color, and 1 point for two green colors 1.5 points for other cases of greening.

(4) Scoring criteria for the aesthetics of architectural sketch space landscape color: 0.5 points for a single color of architectural sketches, 1 point for two kinds of architectural sketches, 1.5 points for other cases of architectural sketches; 0.5 points for single greening colors, and two 1 point for species, 1.5 points for other cases of greening.

Floor pavement aesthetics: mainly judged from the visual effects such as texture, splicing, and layers of the floor pavement.

Through the street view data, using the deep learning technology in the field of artificial intelligence, the raters with professional background first scored 1 to 3 points on the street view data of some parks according to four standards (the specific standards are attached), 1 is low and 2 points Average, 3 points are high, and the park street view data after scoring are used as sample data. In TensorFlow, the second-generation open source artificial intelligence learning system developed by Google, the above park street view data and corresponding score values are called in batches using the Python programming language , through deep learning technology to obtain its nonlinear complex feature selection, so as to have the ability of automatic batch scoring for most of the remaining park street view data, that is, through deep learning model training on part of the park street view data, it is realized that most of the remaining park street view data Data autonomous batch scoring.

evaluation standard:

(1) Scoring criteria for the aesthetics of the ground pavement of the garden road space: 1 point for the use of a single material for the garden road, 1.5 points for the use of two materials for the garden road, and 3 points for the rest of the garden road.

(2) Scoring standard for aesthetics of ground pavement in waterfront space: 1 point for single ground material, 1.5 points for two types of ground materials, and 3 points for other ground materials.

(3) Scoring criteria for the aesthetics of the ground pavement in the square space: 1 point for the use of a single material in the square, 1.5 points for the use of two materials in the square, and 3 points for the rest of the square.

(4) Scoring criteria for the aesthetics of the ground pavement in architectural sketch space: 1 point for a single floor material, 1.5 points for two types of floor materials, and 3 points for other ground materials.

The richness of plant species or landscape elements: it is measured by whether the plant species are diverse or whether the landscape elements are rich. Calculated from street view photo data, the calculation formula is as follows:

V _plant = P _plant / P where V _plant is the richness of plant species or landscape elements, P _plant is the total pixel value of plant species (or landscape elements) in each photo, and P is the total pixel value of each photo.

Waterfront space quality evaluation index content:

Green viewing rate: take the proportion of vegetation as the green viewing rate. Calculated from street view photo data, the calculation formula is as follows:

L=P_vegetation/P In the formula, L is the green viewing rate, P_vegetationis the total pixel value of green vegetation in each photo, and P is the total pixel value of each photo.

Sky Openness: The sky openness is measured by the sky area ratio. Calculated from street view photo data, the calculation formula is as follows:

T= _Psky /P where T is the openness of the sky, Psky is the total pixel value of the sky in each photo, and P is the total pixel value of each photo.

Building perception: Building perception is measured by building ratio. Calculated from street view photo data, the calculation formula is as follows:

B _building = P _building / P where B _building is the perception of the building, P _sidewalk is the total pixel value of the building in each photo, and P is the total pixel value of each photo.

Interface enclosure: The total proportion of buildings, fences, walls, poles, signs and vegetation is taken as the interface enclosure. Calculated from street view photo data, the calculation formula is as follows:

J _interface = P _interface / P where J _interface is the interface enclosure, P _interface is the total pixel value of buildings, fences, walls, poles, signs and vegetation in each photo, and P is the total pixel value of each photo value.

Landscape color aesthetics: It refers to the psychological changes caused by the stimulation of human visual system by the landscape color, and the color factor is considered from the change of the four seasons of the landscape.

Through the street view data, using the deep learning technology in the field of artificial intelligence, the raters with professional background first scored 1 to 3 points on the street view data of some parks according to four standards (the specific standards are attached), 1 is low and 2 points Average, 3 points are high, and the park street view data after scoring are used as sample data. In TensorFlow, the second-generation open source artificial intelligence learning system developed by Google, the above park street view data and corresponding score values are called in batches using the Python programming language , through deep learning technology to obtain its nonlinear complex feature selection, so as to have the ability of automatic batch scoring for most of the remaining park street view data, that is, through deep learning model training on part of the park street view data, it is realized that most of the remaining park street view data Data autonomous batch scoring.

evaluation standard:

(1) Scoring criteria for the color aesthetics of garden road space landscape: 0.5 points for garden roads with a cool color system and a single color, 1 point for a garden road with a neutral color system and a single color, 1.5 points for other garden roads; single green color 0.5 points for green color equal to two, 1 point for other greening conditions, 1.5 points.

(2) Scoring standard for waterfront space landscape color aesthetics: 0.5 points for water body that is black, 1 point for turbid water body that is gray, 1.5 points for clear water body that is colorless and transparent; 0.5 point for single green color, equal to two green colors 1 point for greening and 1.5 points for other cases of greening.

(3) Scoring criteria for the color aesthetics of the square space landscape: 0.5 points for a single square color, 1 point for two square colors, and 1.5 points for the rest of the square; 0.5 points for a single green color, and 1 point for two green colors 1.5 points for other cases of greening.

(4) Scoring criteria for the aesthetics of architectural sketch space landscape color: 0.5 points for a single color of architectural sketches, 1 point for two kinds of architectural sketches, 1.5 points for other cases of architectural sketches; 0.5 points for single greening colors, and two 1 point for species, 1.5 points for other cases of greening.

Floor pavement aesthetics: mainly judged from the visual effects such as texture, splicing, and layers of the floor pavement.

Through the street view data, using the deep learning technology in the field of artificial intelligence, the raters with professional background first scored 1 to 3 points on the street view data of some parks according to four standards (the specific standards are attached), 1 is low and 2 points Average, 3 points are high, and the park street view data after scoring are used as sample data. In TensorFlow, the second-generation open source artificial intelligence learning system developed by Google, the above park street view data and corresponding score values are called in batches using the Python programming language , through deep learning technology to obtain its nonlinear complex feature selection, so as to have the ability to automatically batch score most of the remaining park street view data, that is, through deep learning model training on part of the park street view data, it is realized that most of the remaining park street view data Data autonomous batch scoring.

evaluation standard:

(1) Scoring criteria for the aesthetics of the ground pavement of the garden road space: 1 point for the use of a single material for the garden road, 1.5 points for the use of two materials for the garden road, and 3 points for the rest of the garden road.

(2) Scoring standard for aesthetics of ground pavement in waterfront space: 1 point for single ground material, 1.5 points for two types of ground materials, and 3 points for other ground materials.

(3) Scoring criteria for the aesthetics of the ground pavement in the square space: 1 point for the use of a single material in the square, 1.5 points for the use of two materials in the square, and 3 points for the rest of the square.

(4) Scoring criteria for the aesthetics of the ground pavement in architectural sketch space: 1 point for a single floor material, 1.5 points for two types of floor materials, and 3 points for other ground materials.

The richness of plant species or landscape elements: it is measured by whether the plant species are diverse or whether the landscape elements are rich. Calculated from street view photo data, the calculation formula is as follows:

V _plant = P _plant / P where V _plant is the richness of plant species or landscape elements, P _plant is the total pixel value of plant species (or landscape elements) in each photo, and P is the total pixel value of each photo.

Plaza space quality evaluation index content:

Green viewing rate: take the proportion of vegetation as the green viewing rate. Calculated from street view photo data, the calculation formula is as follows:

L=P_vegetation/P In the formula, L is the green viewing rate, P_vegetationis the total pixel value of green vegetation in each photo, and P is the total pixel value of each photo.

Building perception: Building perception is measured by building ratio. Calculated from street view photo data, the calculation formula is as follows:

B _building = P _building / P where B _building is the perception of the building, P _sidewalk is the total pixel value of the building in each photo, and P is the total pixel value of each photo.

Interface enclosure: The total proportion of buildings, fences, walls, poles, signs and vegetation is taken as the interface enclosure. Calculated from street view photo data, the calculation formula is as follows:

J _interface = P _interface /P formula where J _interface is the interface enclosure, P _interface is the total pixel value of buildings, fences, walls, poles, signs and vegetation in each photo, and P is the total pixel value of each photo value.

Landscape color aesthetics: Refers to the psychological changes caused by the stimulation of the human visual system by the landscape color, considering the color factor from the change of the four seasons of the landscape.

Through the street view data, using the deep learning technology in the field of artificial intelligence, the raters with professional background first scored 1 to 3 points on the street view data of some parks according to four standards (the specific standards are attached), 1 is low and 2 points Average, 3 points are high, and the park street view data after scoring are used as sample data. In TensorFlow, the second-generation open source artificial intelligence learning system developed by Google, the above park street view data and corresponding score values are called in batches using the Python programming language , through deep learning technology to obtain its nonlinear complex feature selection, so as to have the ability of automatic batch scoring for most of the remaining park street view data, that is, through deep learning model training on part of the park street view data, it is realized that most of the remaining park street view data Data autonomous batch scoring.

evaluation standard:

(1) Scoring criteria for the color aesthetics of garden road space landscape: 0.5 points for garden roads with a cool color system and a single color, 1 point for a garden road with a neutral color system and a single color, 1.5 points for other garden roads; single green color 0.5 points for green color equal to two, 1 point for other greening conditions, 1.5 points.

(2) Scoring standard for waterfront space landscape color aesthetics: 0.5 points for water body that is black, 1 point for turbid water body that is gray, 1.5 points for clear water body that is colorless and transparent; 0.5 point for single green color, equal to two green colors 1 point for greening and 1.5 points for other cases of greening.

(3) Scoring criteria for the color aesthetics of the square space landscape: 0.5 points for a single square color, 1 point for two square colors, and 1.5 points for the rest of the square; 0.5 points for a single green color, and 1 point for two green colors 1.5 points for other cases of greening.

(4) Scoring criteria for the aesthetics of architectural sketch space landscape color: 0.5 points for a single color of architectural sketches, 1 point for two kinds of architectural sketches, 1.5 points for other cases of architectural sketches; 0.5 points for single greening colors, and two 1 point for species, 1.5 points for other cases of greening.

Floor pavement aesthetics: mainly judged from the visual effects such as texture, splicing, and layers of the floor pavement.

Through the street view data, using the deep learning technology in the field of artificial intelligence, the raters with professional background first scored 1 to 3 points on the street view data of some parks according to four standards (the specific standards are attached), 1 is low and 2 points Average, 3 points are high, and the park street view data after scoring are used as sample data. In TensorFlow, the second-generation open source artificial intelligence learning system developed by Google, the above park street view data and corresponding score values are called in batches using the Python programming language , through deep learning technology to obtain its nonlinear complex feature selection, so as to have the ability of automatic batch scoring for most of the remaining park street view data, that is, through deep learning model training on part of the park street view data, it is realized that most of the remaining park street view data Data autonomous batch scoring.

evaluation standard:

(1) Scoring criteria for the aesthetics of the ground pavement of the garden road space: 1 point for the use of a single material for the garden road, 1.5 points for the use of two materials for the garden road, and 3 points for the rest of the garden road.

(2) Scoring standard for aesthetics of ground pavement in waterfront space: 1 point for single ground material, 1.5 points for two types of ground materials, and 3 points for other ground materials.

(3) Scoring criteria for the aesthetics of the ground pavement in the square space: 1 point for the use of a single material in the square, 1.5 points for the use of two materials in the square, and 3 points for the rest of the square.

(4) Scoring criteria for the aesthetics of the ground pavement in architectural sketch space: 1 point for a single floor material, 1.5 points for two types of floor materials, and 3 points for other ground materials.

The richness of plant species or landscape elements: it is measured by whether the plant species are diverse or whether the landscape elements are rich. Calculated from street view photo data, the calculation formula is as follows:

V _plants = P _plants / P where V _plants are the richness of plant species or landscape elements, P _plants are the total pixel values of plant species (or landscape elements) in each photo, and P is the total pixel value of each photo.

Contents of space quality evaluation indicators for architectural sketches:

Green viewing rate: take the proportion of vegetation as the green viewing rate. Calculated from street view photo data, the calculation formula is as follows:

L=P _vegetation /P where L is the green viewing rate, P _vegetation is the total pixel value of green vegetation in each photo, and P is the total pixel value of each photo.

Landscape color aesthetics: Refers to the psychological changes caused by the stimulation of the human visual system by the landscape color, considering the color factor from the change of the four seasons of the landscape.

Through the street view data, using the deep learning technology in the field of artificial intelligence, the raters with professional background first scored 1 to 3 points on the street view data of some parks according to four standards (the specific standards are attached), 1 is low and 2 points Average, 3 points are high, and the park street view data after scoring are used as sample data. In TensorFlow, the second-generation open source artificial intelligence learning system developed by Google, the above park street view data and corresponding score values are called in batches using the Python programming language , through deep learning technology to obtain its nonlinear complex feature selection, so as to have the ability of automatic batch scoring for most of the remaining park street view data, that is, through deep learning model training on part of the park street view data, it is realized that most of the remaining park street view data Data autonomous batch scoring.

evaluation standard:

(1) Scoring criteria for the color aesthetics of garden road space landscape: 0.5 points for garden roads with a cool color system and a single color, 1 point for a garden road with a neutral color system and a single color, 1.5 points for other garden roads; single green color 0.5 points for green color equal to two, 1 point for other greening conditions, 1.5 points.

(2) Scoring standard for waterfront space landscape color aesthetics: 0.5 points for water body that is black, 1 point for turbid water body that is gray, 1.5 points for clear water body that is colorless and transparent; 0.5 point for single green color, equal to two green colors 1 point for greening and 1.5 points for other cases of greening.

(3) Scoring criteria for the color aesthetics of the square space landscape: 0.5 points for a single square color, 1 point for two square colors, and 1.5 points for the rest of the square; 0.5 points for a single green color, and 1 point for two green colors 1.5 points for other cases of greening.

(4) Scoring criteria for the aesthetics of architectural sketch space landscape color: 0.5 points for a single color of architectural sketches, 1 point for two kinds of architectural sketches, 1.5 points for other cases of architectural sketches; 0.5 points for single greening colors, and two 1 point for species, 1.5 points for other cases of greening.

Floor pavement aesthetics: mainly judged from the visual effects such as texture, splicing, and layers of the floor pavement.

Through the street view data, using the deep learning technology in the field of artificial intelligence, the raters with professional background first scored 1 to 3 points on the street view data of some parks according to four standards (the specific standards are attached), 1 is low and 2 points Average, 3 points are high, and the park street view data after scoring are used as sample data. In TensorFlow, the second-generation open source artificial intelligence learning system developed by Google, the above park street view data and corresponding score values are called in batches using the Python programming language , through deep learning technology to obtain its nonlinear complex feature selection, so as to have the ability of automatic batch scoring for most of the remaining park street view data, that is, through deep learning model training on part of the park street view data, it is realized that most of the remaining park street view data Data autonomous batch scoring.

evaluation standard:

(1) Scoring criteria for the aesthetics of the ground pavement of the garden road space: 1 point for the use of a single material for the garden road, 1.5 points for the use of two materials for the garden road, and 3 points for the rest of the garden road.

(2) Scoring standard for aesthetics of ground pavement in waterfront space: 1 point for single ground material, 1.5 points for two types of ground materials, and 3 points for other ground materials.

(3) Scoring criteria for the aesthetics of the ground pavement in the square space: 1 point for the use of a single material in the square, 1.5 points for the use of two materials in the square, and 3 points for the rest of the square.

(4) Scoring criteria for the aesthetics of the ground pavement in architectural sketch space: 1 point for a single floor material, 1.5 points for two types of floor materials, and 3 points for other ground materials.

The richness of plant species or landscape elements: it is measured by whether the plant species are diverse or whether the landscape elements are rich. Calculated from street view photo data, the calculation formula is as follows:

V _plants = P _plants / P where V _plants are the richness of plant species or landscape elements, P _plants are the total pixel values of plant species (or landscape elements) in each photo, and P is the total pixel value of each photo.

The weight values of each index in the evaluation index system of the present invention are obtained by using the AHP and Delphi methods, specifically using YAAHP software to construct the levels of each index in the evaluation index system, and then comparing the importance of the indexes in pairs , assign the corresponding value, and pass the consistency test. After repeating 20 times, the average weight value of each index in the evaluation index system is obtained.

The evaluation index system and weight are shown in the table below:

Table 1: Comprehensive evaluation index system and weight table of urban park space quality;

Embodiment 4. On the basis of Embodiment 3, the construction of the comprehensive evaluation index system of urban park space quality is specifically: through the construction of the park physical environment space quality evaluation index system and the construction of the park perception environment space quality evaluation index system, finally formed The comprehensive evaluation index system of urban park space quality, and the weight value of each index is determined by the analytic hierarchy process.

It will be apparent to those skilled in the art that the invention is not limited to the details of the above-described exemplary embodiments, but that the invention can be embodied in other specific forms without departing from the spirit or essential characteristics of the invention. Accordingly, the embodiments should be regarded in all points of view as exemplary and not restrictive, the scope of the invention being defined by the appended claims rather than the foregoing description, and it is therefore intended that the scope of the invention be defined by the appended claims rather than by the foregoing description. All changes within the meaning and range of equivalents of the elements are embraced in the present invention. Any reference sign in a claim should not be construed as limiting the claim concerned.

In addition, it should be understood that although this specification is described according to implementation modes, not each implementation mode only contains an independent technical solution, and this description in the specification is only for clarity, and those skilled in the art should take the specification as a whole , the technical solutions in the various embodiments can also be properly combined to form other implementations that can be understood by those skilled in the art.

Claims (8)

1. A method for constructing a comprehensive evaluation index system of urban park space quality is characterized by comprising the following steps;

dividing the urban park space quality into two dimensions of a material environment and a perception environment, and two dimensions of an external community and an internal community;

constructing an urban park material environment space quality evaluation index system based on the two dimensions and the two scales;

constructing a perception environment space quality evaluation index system of the city park based on the two dimensions and the two scales;

and determining the weight of the index in each dimension and each scale by adopting an analytic hierarchy process, and finally constructing a comprehensive evaluation index system for the space quality of the city park.

2. The method as claimed in claim 1, wherein the step of constructing the city park physical environment space quality evaluation index system specifically comprises: the method comprises the steps of constructing external community scale indexes in the park material environment and constructing internal scale indexes in the park material environment.

3. The method as claimed in claim 2, wherein the building of external community scale indexes in the park material environment comprises: selecting an external community scale evaluation index, wherein the external community scale evaluation index comprises the following steps: standardizing angle integration degree, bus stop density, distance from a subway entrance, function density, function mixing degree, land property of peripheral land parcels and development strength of the peripheral land parcels, and quantifying selected indexes to obtain external community scale indexes in the park material environment.

4. The method as claimed in claim 2, wherein the step of constructing the internal dimension index under the park material environment comprises: selecting an internal scale evaluation index, wherein the internal scale evaluation index comprises the following steps: standardizing angle integration degree, level of matched service facilities, functional diversity and water cleanliness, and quantizing indexes by adopting a space syntax and field investigation method to obtain internal scale indexes of the park in a material environment.

5. The method as claimed in claim 1, wherein the step of constructing the city park perception environment space quality evaluation index system comprises: the method comprises the steps of building an external community scale index in a park perception environment and building an internal scale index in the park perception environment.

6. The method as claimed in claim 5, wherein the building of external community scale indexes in the park-aware environment comprises: and selecting an external community scale evaluation index, wherein the external community scale evaluation index comprises a green visibility rate, a sky openness degree, an interface surrounding degree, a motorization degree, a sidewalk perception degree, a building perception degree and a human flow ratio, and obtaining the external community scale index in the park perception environment.

7. The method as claimed in claim 5, wherein the building of the internal dimension index under the park perception environment comprises: subdividing the park interior space into: 4 types of garden path space, waterfront space, square space and building small article space are used for building a park perception environment space quality evaluation index system in a more refined manner, and internal scale evaluation indexes are selected and comprise: and obtaining the internal dimension index of the park in the perception environment.

8. The method for constructing an index system for the comprehensive evaluation of the spatial quality of urban parks according to any one of claims 1 to 7, wherein the weighted values of the indexes in the index system are obtained by using AHP (analytic hierarchy process) and Delphi method, specifically YAAHP (Yaahp software) to construct the index levels in the index system, and the index importance is compared with each other to give corresponding values, and the consistency test is carried out, and the average weighted values of the indexes in the index system are obtained after repeating for 20 times.

Images