Method and Apparatus for Analysis of Thermal Environment Characteristics by Physical Spatial Type in Urban Areas Based on UAV

Abstract

A method and apparatus for analyzing thermal environment characteristics by physical space type in urban areas based on unmanned aerial vehicles are presented. The method for analyzing thermal environment characteristics by physical space type in urban areas based on unmanned aerial vehicles proposed in this invention includes the steps of classifying physical environment types according to unmanned aerial vehicle-based NDVI (Normalized Different Vegetation Index) and SVF (Sky View Factor) characteristics for a target site for thermal environment characteristic analysis, the step of analyzing thermal comfort characteristics of the target site using microclimate modeling (ENVI-met), and the step of analyzing the distribution of unmanned aerial vehicle-based thermal comfort by comparing the classified physical environment types and the analyzed thermal comfort characteristics.

Description

Technology Field

[0001] The present invention relates to a method and apparatus for analyzing thermal environment characteristics by physical space type in urban areas based on an unmanned aerial vehicle. Background Technology

[0002] Changes in the physical environment, such as the increase in artificial urban covering materials, are acting as a significant cause of urban climate change problems, including recurring heatwaves and urban heat islands every summer.

[0003] In order to identify the causes of urban thermal environment problems and establish continuous improvement measures, it is necessary to understand the characteristics of the thermal environment by considering precise and diverse physical environmental factors of urban spaces.

[0004] While field measurements, satellite imagery-based remote sensing, and microclimate modeling are utilized to identify thermal environmental characteristics considering the physical environment of urban areas, the use of satellite imagery is limited to analyzing thermal environmental characteristics based on precise physical environmental factors in urban areas due to the predominantly medium to low resolution.

[0005] When utilizing climate modeling, there are limitations in accurately predicting the thermal environment due to the difficulty in constructing input data that closely resembles the actual space, and field measurements have limitations in analyzing the thermal environment of extensive urban areas.

[0006] With the recent advancement of unmanned aerial vehicle and sensor technologies, they are being effectively utilized to identify precise physical spatial factors in urban areas.

[0007] In this regard, while numerous studies based on high-resolution unmanned aerial vehicle (UAV) imagery are being conducted in the field of urban thermal environments, research identifying the thermal environmental characteristics of urban areas by considering precise physical spaces based on UAVs is lacking. In particular, there is a current absence of research identifying the characteristics of thermal comfort experienced by humans in actual outdoor spaces.

[0008] Therefore, it is necessary to develop technology to identify the distribution of thermal comfort according to physical space types by utilizing high-resolution unmanned aerial vehicle imagery to characterize physical spaces and climate modeling to analyze thermal comfort. Prior art literature

[0009] An, SM, Son, HG, Lee, KS and Yi, CY 2016. A study of the urban tree canopy mean radiant temperature mitigation estimation. Journal of KILA 44(1):93-106 (Ahn, Seung-Man, Son, Hak-Ki, Lee, Kyu-Suk, and Yi, Chae-Yeon. 2016. A study on the estimation of mean radiant temperature reduction in urban forests during summer. Journal of the Korean Society of Landscape Architects 44(1):93-106). Song, BG and Park, KH 2019. A study on the relationship between land cover type and urban temperature - focused on Gimhae city -. Journal of the Korean Association of Geographic Information Studies 22(2):65-81 (Song, Bong-Geun, and Park, Kyoung-Hoon. 2019. Analysis of the relationship between land cover type characteristics and urban temperature - focused on Gimhae city -. Journal of the Korean Association of Geographic Information Studies 22(2): 65-81). The problem to be solved

[0010] The technical problem that the present invention aims to solve is to provide a method and apparatus for identifying the distribution of thermal comfort according to physical space types by identifying physical space characteristics using high-resolution unmanned aerial vehicle imagery data and performing thermal comfort analysis using microclimate modeling. means of solving the problem

[0011] In one aspect, the unmanned aerial vehicle-based method for analyzing thermal environment characteristics by physical space type in urban areas proposed in the present invention includes the steps of classifying physical environment types according to unmanned aerial vehicle-based NDVI (Normalized Different Vegetation Index) and SVF (Sky View Factor) characteristics for a target site for thermal environment characteristic analysis, the step of analyzing thermal comfort characteristics of the target site using microclimate modeling (ENVI-met), and the step of analyzing the unmanned aerial vehicle-based thermal comfort distribution by comparing the classified physical environment types and the analyzed thermal comfort characteristics.

[0012] The step of classifying physical environment types based on NDVI (Normalized Different Vegetation Index) and SVF (Sky View Factor) characteristics for a target site for thermal environment characteristic analysis according to an embodiment of the present invention involves collecting unmanned aerial vehicle images by capturing multispectral images and optical images through an unmanned aerial vehicle to classify physical environment types.

[0013] According to an embodiment of the present invention, the RGB and multispectral images of the collected unmanned aerial vehicle images are produced into orthophotos using image joining software, and a Digital Surface Model (DSM) is produced using optical images for SVF analysis.

[0014] The step of classifying the physical environment type according to the unmanned aerial vehicle-based NDVI (Normalized Different Vegetation Index) and SVF (Sky View Factor) characteristics for a target site for thermal environment characteristic analysis according to an embodiment of the present invention calculates the NDVI and SVF characteristics based on the produced unmanned aerial vehicle image, and utilizes the land use and cover type and the degree of density of buildings and trees according to the NDVI and SVF characteristics as classification factors for the physical environment type of the target site.

[0015] According to an embodiment of the present invention, NDVI is a unitless radiative value calculated through reflection values ​​in the near-infrared and red band regions, and SVF is an indicator representing the degree of building density by analyzing the ratio of the area where the sky is visible at a surface observation point, requiring urban morphological data, and is calculated through a terrain analysis function utilizing a DSM based on high-resolution RGB images of an unmanned aerial vehicle, and for physical environment type classification, the number of physical environment type clusters is derived through Python-based inertia analysis, and then K-means clustering is performed to classify the physical environment type.

[0016] The step of analyzing the thermal comfort characteristics of the target site using microclimate modeling (ENVI-met) according to an embodiment of the present invention comprises a thermal comfort factor T for measuring and expressing radiant energy affecting the thermal sensation of the human body among the simulation results using microclimate modeling (ENVI-met). MRT Data is extracted and converted into raster data to analyze the thermal comfort characteristics of the target site.

[0017] The step of analyzing the thermal comfort characteristics of the target site using microclimate modeling (ENVI-met) according to an embodiment of the present invention comprises constructing an input file (.IN) for physical environmental factors in the microclimate modeling simulation by inputting building shape, height, vegetation height and species, and covering material through RGB optical images of an unmanned aerial vehicle and field surveys, and constructing a simulation condition file (.CF) by setting coordinates, grid size, initial wind speed and direction, temperature, and simulation time.

[0018] The step of analyzing the unmanned aerial vehicle-based thermal comfort distribution by comparing classified physical environment types and the analyzed thermal comfort characteristics according to an embodiment of the present invention comprises thermal comfort (T) extracted from a simulation using microclimate modeling (ENVI-met).MRT Using ), a distribution map of the error between the modeled values ​​and the predicted values ​​is constructed, and thermal comfort (T) by physical environment type and time period is MRT ), hourly thermal comfort according to NDVI and SVF (T MRT Analyzes the distribution.

[0019] In another aspect, the unmanned aerial vehicle-based thermal environment characteristic analysis device for each physical space type in an urban area proposed in the present invention includes a physical environment type classification unit that classifies physical environment types according to unmanned aerial vehicle-based NDVI (Normalized Different Vegetation Index) and SVF (Sky View Factor) characteristics for a target site for thermal environment characteristic analysis, a thermal comfort characteristic analysis unit that analyzes thermal comfort characteristics of the target site using microclimate modeling (ENVI-met), and a thermal comfort distribution analysis unit that analyzes the unmanned aerial vehicle-based thermal comfort distribution by comparing the classified physical environment types and the analyzed thermal comfort characteristics. Effects of the invention

[0020] Through the method and apparatus for analyzing thermal environment characteristics by physical space type in urban areas based on unmanned aerial vehicles according to the embodiments of the present invention, it is possible to formulate plans considering thermal comfort for open areas such as plazas and parks with high population flow in urban areas, and it can be used to quickly and accurately identify the causes of heatwaves and tropical nights that occur repeatedly every summer, and it can be used as important basic data for national disaster and safety management. Brief explanation of the drawing

[0021] FIG. 1 is a flowchart illustrating a method for analyzing thermal environment characteristics by physical space type in an urban area based on an unmanned aerial vehicle according to an embodiment of the present invention. FIG. 2 is a diagram showing the configuration of a thermal environment characteristic analysis device based on an unmanned aerial vehicle and a type of physical space in an urban area according to an embodiment of the present invention. FIG. 3 is a drawing showing a target site for analyzing thermal environment characteristics by physical space type according to an embodiment of the present invention. FIG. 4 is a diagram showing an IN file of ENVI-met modeling according to one embodiment of the present invention. Figure 5 is a diagram showing the results of producing NDVI and SVF for identifying physical factors using UAV images according to one embodiment of the present invention. Figure 6 is a graph showing the results of analyzing Python-based inertia according to one embodiment of the present invention. FIG. 7 is a diagram showing the K-means cluster analysis results according to one embodiment of the present invention. FIG. 8 is a diagram showing box plots of NDVI and SVF by cluster type according to one embodiment of the present invention. FIG. 9 is a diagram showing a cross-sectional schematic diagram by cluster type according to one embodiment of the present invention. FIG. 10 shows T within the park, a research site utilizing NVI-met according to an embodiment of the present invention. MRT This is a drawing showing the modeling results. FIG. 11 shows a physical space type and T by time period according to an embodiment of the present invention. MRT This is a diagram showing the results of the characteristic analysis. FIG. 12 shows T over time according to physical spatial factors NDVI and SVF according to an embodiment of the present invention. MR This is a diagram showing the results of analyzing the T distribution. FIG. 13 is a diagram showing the location of a point of low density in kernel density according to one embodiment of the present invention. Specific details for implementing the invention

[0022] The increase in the area of ​​urbanized regions due to urbanization has transformed existing natural spaces, such as forests and grasslands, into artificial covering materials like asphalt and concrete. As a result, the urban heat island effect, in which summer temperatures in urban areas are more than 2°C higher than in suburban areas, occurs frequently, and citizens residing in urban areas are currently exposed to various urban thermal environmental problems, such as heatwaves and tropical nights. In large cities with high development densities, various measures are being attempted to continuously improve urban heat island and thermal environmental issues, including identifying the causes of such problems and establishing policy and institutional foundations based on this analysis.

[0023] To date, various studies have been conducted both domestically and internationally to identify the causes of urban thermal environment problems and establish mitigation measures. Prior research has utilized diverse analytical tools, such as field measurements, satellite imagery-based remote sensing, and microclimate modeling. These studies cover the interrelationship between urban physical environmental factors and the climate environment, the classification of spatial types considering urban climate formation factors, and research on improving urban thermal comfort using climate modeling. However, while it is necessary to identify the causes of urban thermal environment problems and develop sustainable improvement measures by understanding thermal environmental characteristics that consider precise and diverse physical environmental factors of urban spaces, prior research has limitations in this regard. In the case of studies utilizing satellite imagery, most are based on medium to low resolution, which limits the analysis of thermal environmental characteristics based on precise physical environmental factors in urban areas. Furthermore, studies using climate modeling face limitations in accurately predicting the thermal environment due to constraints in constructing input data that closely resembles actual space; additionally, they suffer from the disadvantage that predicting thermal comfort requires excessive time as the target area and modeling time increase. Furthermore, while field measurement-based research can consider precise physical environments, it has limitations in analyzing the thermal environment of extensive urban areas because the analysis is restricted to the vicinity of the measurement point.

[0024] Meanwhile, with the recent advancements in Unmanned Aerial Vehicle (UAV) and sensor technologies, they are being effectively utilized to identify precise physical environmental factors (in other words, spatial factors) in urban areas. This can overcome the limitations of using existing low-to-medium resolution satellite imagery. Furthermore, UAVs offer the advantage of overcoming constraints regarding time and location, allowing researchers to easily acquire information on specific sites at their desired times. Leveraging these advantages, these technologies are being applied in various fields such as disaster prevention, surveying, and broadcasting, and numerous studies based on high-resolution UAV imagery are also being conducted in the field of urban thermal environments. However, technologies that identify the thermal environmental characteristics of urban areas by considering precise physical space based on UAVs are currently lacking; in particular, technologies that identify the characteristics of thermal comfort experienced by people in actual outdoor spaces are absent. To improve the urban thermal environment, it is crucial to precisely determine the thermal comfort actually felt by people based on urban physical spatial factors, which necessitates the development of urban spatial planning strategies.

[0025] Accordingly, in an embodiment of the present invention, thermal comfort analysis was performed using climate modeling and classification of physical spatial factors and types utilizing high-resolution UAV imagery data targeting Changwon City, Gyeongsangnam-do, to derive thermal comfort characteristics according to physical spatial types, and measures for improving urban thermal comfort were proposed. Hereinafter, an embodiment of the present invention will be described in detail with reference to the attached drawings.

[0027] FIG. 1 is a flowchart illustrating a method for analyzing thermal environment characteristics by physical space type in an urban area based on an unmanned aerial vehicle according to an embodiment of the present invention.

[0028] The method for analyzing thermal environment characteristics by physical space type in an urban area based on an unmanned aerial vehicle proposed in the present invention includes the step (110) of classifying physical environment types according to the characteristics of the unmanned aerial vehicle-based NDVI (Normalized Different Vegetation Index) and SVF (Sky View Factor) for a target area for thermal environment characteristic analysis, the step (120) of analyzing thermal comfort characteristics of the target area using microclimate modeling (ENVI-met), and the step (130) of analyzing the distribution of thermal comfort based on an unmanned aerial vehicle by comparing the classified physical environment types and the analyzed thermal comfort characteristics.

[0029] In step (110), the physical environment type is classified according to the NDVI (Normalized Different Vegetation Index) and SVF (Sky View Factor) characteristics based on unmanned aerial vehicle for the analysis of thermal environment characteristics.

[0030] First, unmanned aerial vehicle images (RGB, multispectral, etc.) are acquired (111), and at this time, to classify the physical environment type, multispectral images and optical images are taken through the unmanned aerial vehicle to collect unmanned aerial vehicle images.

[0031] According to an embodiment of the present invention, the RGB and multispectral images of the collected unmanned aerial vehicle images are produced into orthophotos using image joining software, and a Digital Surface Model (DSM) is produced using optical images for SVF analysis.

[0032] Next, an analysis of the physical environment characteristics of the target site is performed (112), and then a classification of types by physical characteristics is performed (113).

[0033] In the step (110) of classifying the physical environment type according to the unmanned aerial vehicle-based NDVI (Normalized Different Vegetation Index) and SVF (Sky View Factor) characteristics for a target site for thermal environment characteristic analysis according to an embodiment of the present invention, the NDVI and SVF characteristics are calculated based on the produced unmanned aerial vehicle image, and the land use and cover type and the degree of density of buildings and trees according to the NDVI and SVF characteristics are used as classification factors for the physical environment type of the target site.

[0034] According to an embodiment of the present invention, NDVI is a unitless radiative value calculated through reflection values ​​in the near-infrared and red band regions, and SVF is an indicator representing the degree of building density by analyzing the ratio of the area where the sky is visible at a surface observation point, requiring urban morphological data, and is calculated through a terrain analysis function utilizing a DSM based on high-resolution RGB images of an unmanned aerial vehicle, and for physical environment type classification, the number of physical environment type clusters is derived through Python-based inertia analysis, and then K-means clustering is performed to classify the physical environment type.

[0035] In step (120), the thermal comfort characteristics of the target area are analyzed using microclimate modeling (ENVI-met). A simulation is performed using microclimate modeling (ENVI-met) according to an embodiment of the present invention (121), and among the simulation results, thermal comfort (T) for measuring and expressing radiant energy affecting the thermal sense of the human body is MRT ) data is extracted and converted into raster data to analyze the thermal comfort characteristics of the target area (122).

[0036] In the step (120) of analyzing the thermal comfort characteristics of the target site using microclimate modeling (ENVI-met) according to an embodiment of the present invention, the input file (.IN) for physical environmental factors in the microclimate modeling simulation is constructed by inputting building shape, height, vegetation height and species, and covering material through RGB optical images of an unmanned aerial vehicle and field surveys, and the simulation condition file (.CF) is constructed by setting coordinates, grid size, initial wind speed and direction, and temperature.

[0037] In step (130), the unmanned aerial vehicle-based thermal comfort distribution is analyzed by comparing the classified physical environment types and the analyzed thermal comfort characteristics. The thermal comfort (T) extracted from a simulation using microclimate modeling (ENVI-met) according to an embodiment of the present invention MRT Using ), a distribution map of the error between the modeled value and the predicted value was constructed and compared (131), and the thermal comfort (T) by physical environment type and time of day was analyzed. MRT ), hourly thermal comfort (T MRT Analyzes the distribution (132).

[0039] FIG. 2 is a diagram showing the configuration of a thermal environment characteristic analysis device based on an unmanned aerial vehicle and a type of physical space in an urban area according to an embodiment of the present invention.

[0040] The unmanned aerial vehicle-based thermal environment characteristic analysis device (200) for each physical space type in an urban area proposed in the present invention includes a physical environment type classification unit (210), a thermal comfort characteristic analysis unit (220), and a thermal comfort distribution analysis unit (230).

[0041] A physical environment type classification unit (210) according to an embodiment of the present invention classifies the physical environment type for a target site for thermal environment characteristic analysis based on the unmanned aerial vehicle-based NDVI (Normalized Different Vegetation Index) and SVF (Sky View Factor) characteristics.

[0042] First, unmanned aerial vehicle (UAV) images (RGB, multispectral, etc.) are acquired. At this time, to classify the type of physical environment, multispectral and optical images are captured via the UAV to collect the UAV images.

[0043] According to an embodiment of the present invention, the RGB and multispectral images of the collected unmanned aerial vehicle images are produced into orthophotos using image joining software, and a Digital Surface Model (DSM) is produced using optical images for SVF analysis.

[0044] Next, the physical environment type classification unit (210) according to an embodiment of the present invention performs an analysis of physical environment characteristics of the target site and then performs a type classification according to physical characteristics.

[0045] A physical environment type classification unit (210) according to an embodiment of the present invention calculates NDVI and SVF characteristics based on the produced unmanned aerial vehicle image, and uses land use and cover type and the degree of density of buildings and trees according to the NDVI and SVF characteristics as physical environment type classification factors of the target site.

[0046] According to an embodiment of the present invention, NDVI is a unitless radiative value calculated through reflection values ​​in the near-infrared and red band regions, and SVF is an indicator representing the degree of building density by analyzing the ratio of the area where the sky is visible at a surface observation point, requiring urban morphological data, and is calculated through a terrain analysis function utilizing a DSM based on high-resolution RGB images of an unmanned aerial vehicle, and for physical environment type classification, the number of physical environment type clusters is derived through Python-based inertia analysis, and then K-means clustering is performed to classify the physical environment type.

[0047] The thermal comfort characteristic analysis unit (220) according to an embodiment of the present invention analyzes the thermal comfort characteristics of the target site by utilizing microclimate modeling (ENVI-met). A simulation is performed using microclimate modeling (ENVI-met) according to an embodiment of the present invention, and among the simulation results, thermal comfort (T) for measuring and expressing radiant energy affecting the thermal sense of the human body is MRT ) data is extracted and converted into raster data to analyze the thermal comfort characteristics of the target area.

[0048] The thermal comfort characteristic analysis unit (220) according to an embodiment of the present invention constructs the input file (.IN) for physical environmental factors in the microclimate modeling simulation by inputting building shape, height, vegetation height and species, and covering material through RGB optical images of an unmanned aerial vehicle and field surveys, and constructs the simulation condition file (.CF) by setting coordinates, grid size, initial wind speed and wind direction, and temperature.

[0049] The thermal comfort distribution analysis unit (230) according to an embodiment of the present invention analyzes the unmanned aerial vehicle-based thermal comfort distribution by comparing the classified physical environment types and the analyzed thermal comfort characteristics. The thermal comfort (T) extracted from a simulation using microclimate modeling (ENVI-met) according to an embodiment of the present invention MRT Using ), a distribution map of the error between modeled and predicted values ​​was constructed and compared, and thermal comfort (T by physical environment type and time of day) MRT ), hourly thermal comfort (T MRT Analyzes the distribution.

[0051] FIG. 3 is a drawing showing a target site for analyzing thermal environment characteristics by physical space type according to an embodiment of the present invention.

[0052] In an embodiment of the present invention, as illustrated in FIG. 3, an analysis of thermal environment characteristics by physical space type was performed on Yongji Cultural Park (35°13'53N, 128°41'05"E), located within the downtown area of ​​Seongsan-gu, Changwon-si, Gyeongsangnam-do. Changwon-si, where the subject site according to the embodiment of the present invention is located, is a region with very high temperatures, with average summer temperatures exceeding 25°C from 2013 to 2019. It is a basin-shaped topography surrounded by mountains with an altitude of over 500m, such as Jangboksan, Bulmosan, and Jeongbyeongsan. Furthermore, due to the characteristics of a planned city, the urban area is distinctly separated, and the urban heat island phenomenon occurs frequently during the summer due to these spatial characteristics. The subject site, Yongji Cultural Park, has an area of ​​approximately 80,000㎡ and is separated by a 10-lane road. Inside the park, trees less than approximately 3m tall are planted along the edges, while grassland is distributed in the center. Park (A), located on the left side of Yongji Cultural Park, has a gravel pedestrian walkway It is covered. The park (B) located on the right has a parking lot and a performance stage installed to the south, and is covered with various ground covering materials such as asphalt, paving blocks, and marble. In addition, to the north, there are resting areas such as sculptures, a small playground with a urethane floor, and benches. Around the park, government offices are located to the top and bottom, a lake park to the left, and a three-story detached house building to the right.

[0053] In this invention, multispectral and optical images of a UAV were captured to analyze the physical space type. According to an embodiment of this invention, the shooting time was conducted between 12:00 and 14:00 on August 21, 2020, and the weather conditions at the time of measurement were clear, with a wind speed of 0.8 m / s and a temperature of 31.1°C. For multispectral image capture, a Rededge M multispectral camera mounted on a 3DR Solo from 3DR was used. In the case of the 3DR Solo, the calibration panel was photographed before flight for radiation correction, and the shooting was performed at an altitude of 100 m (GSD: 7.4 cm / px) and a overlap of 75%. For optical image capture, a Zenmuse X3 camera was used on a DJI Inspire 1 airframe, and images were captured at an altitude of 150 m (GSD: 6.5 cm / px) and a overlap of 90%. The collected RGB and multispectral images were converted into orthophotos using PIX4D mapper S / W, and a Digital Surface Model (DSM) was constructed using optical images for SVF analysis.

[0054] Physical spatial types affecting thermal comfort in urban areas include land use and land cover types, as well as the density and arrangement of buildings and trees. Accordingly, in this invention, the Normalized Differential Vegetation Index (NDVI) and Sky View Factor (SVF) were generated based on produced unmanned aerial vehicle (UAV) imagery, and the land use and land cover types, along with the degree of building and tree density, were utilized as classification factors for the physical spatial types of the target site. The NDVI is a unitless radiation value ranging from -1 to 1, calculated using reflection values ​​in the near-infrared and red band regions, and can be generated using Equation 1 below:

[0055] (1)

[0056] Next, SVF requires urban morphology data as an indicator representing the degree of building density by analyzing the ratio of the area where the sky is visible from a surface observation point. In this invention, SVF was produced using the terrain analysis function of SAGA GIS S / W based on UAV high-resolution RGB image-based DSM (e.g., spatial resolution 6.8 cm / px) as basic data.

[0057] Here, multispectral images (e.g., spatial resolution 6.8 cm / px) captured in the near-infrared region (820–860 nm) for NIR and the red region (663–673 nm) for Red were utilized. Since the produced NDVI and SVF have a high resolution of 6.8 cm / px, there are limitations in comparing them with ENVI-met modeling results analyzed at 2×2 m; therefore, in this invention, the spatial resolution of the images used for physical space type classification was converted to 2 m. For type classification, Python-based inertia analysis was performed to derive the appropriate number of physical space type clusters, and then K-means clustering was performed to classify physical environment types.

[0059] FIG. 4 is a diagram showing an IN file of ENVI-met modeling according to one embodiment of the present invention.

[0060] In this invention, thermal comfort of the target site according to an embodiment of the invention was analyzed using ENVI-met simulation. The analysis range was set to 430m in width and 230m in length, including the area surrounding the park, and the grid was constructed with 215 grids horizontally, 115 grids vertically, and 30 grids vertically, with a grid spacing of 2m. The input file (.IN), which serves as input data for physical factors in the ENVI-met simulation, was constructed by inputting building shapes and heights, vegetation heights and species, and covering materials through RGB optical images from unmanned aerial vehicles and field surveys. The configuration file (.CF), which serves as the simulation conditions, set coordinates, grid size, initial wind speed and direction, and temperature as shown in Table 1.

[0061]

[0062]

[0063] Here, initial weather data was obtained from a nearby weather station. The modeling period was set to 24 hours, from 05:00 to 05:00 the following day. The start time was set to 05:00 considering the modeling stabilization time according to conventional technology. Among the simulation results, T, which enables objective measurement and representation of radiant energy affecting the thermal sensation of the human body at a height of 1.4m at 2-hour intervals from 10:00 to 16:00 MRT Data was extracted and converted into raster data using ArcGIS Pro software to analyze the thermal comfort characteristics of the target site.

[0065] Figure 5 is a diagram showing the results of producing NDVI and SVF for identifying physical factors using UAV images according to one embodiment of the present invention.

[0066] The results of generating NDVI and SVF to identify physical factors using UAV images according to one embodiment of the present invention are as shown in FIG. 5. FIG. 5(a) is the result of generating NDVI, and FIG. 5(b) is the result of generating SVF.

[0067] According to one embodiment of the present invention, the NDVI values ​​ranged from a minimum of -0.57 to a maximum of 0.95. Areas with dense trees showed the highest NDVI of 0.8 or higher, while grasslands showed NDVI values ​​of approximately 0.6 to 0.8. On the other hand, artificial land cover materials such as asphalt, granite, and marble showed a low NDVI of 0.2 or lower, exhibiting a distinct difference from trees.

[0068] In the case of SVF according to one embodiment of the present invention, values ​​ranging from a minimum of 0 to a maximum of 1 were observed. It was confirmed that in areas where trees are dense and artificial structures such as buildings are adjacent, the SVF was low at 0.4 or lower, while in other open spaces, the SVF was high at 0.8 or higher.

[0070] Figure 6 is a graph showing the results of analyzing Python-based inertia according to one embodiment of the present invention.

[0071] Figure 6 shows the results of a Python-based inertia analysis to determine the number of physical spatial types using NDVI and SVF. Since the spatial resolution of images captured by RGB and multi-spectral sensors differs, the spatial resolution of the UAV images was converted to 2m for use. As a result of the analysis, four clusters with an inertia of 500 or less were determined to be the optimal number of physical spatial types. Therefore, the number of clusters was set to four, and a K-means cluster analysis was performed.

[0073] FIG. 7 is a diagram showing the K-means cluster analysis results according to one embodiment of the present invention.

[0074] FIG. 8 is a diagram showing box plots of NDVI and SVF by cluster type according to one embodiment of the present invention.

[0075] Figures 7 and 8 are the results of K-means cluster analysis according to an embodiment of the present invention.

[0077] FIG. 9 is a diagram showing a cross-sectional schematic diagram by cluster type according to one embodiment of the present invention.

[0078] Figure 9 presents a cross-sectional schematic diagram by cluster type. Cluster 1 was classified as a type located mainly in areas around trees, with an average NDVI value of 0.78 and an average SVF value of 0.87.

[0079] Cluster 2 showed an average NDVI of 0.26 and an average SVF of 0.71, and was identified as a non-vegetated area with a low NDVI corresponding to an open space. Cluster 3 showed an NDVI of 0.82 and an SVF of 0.61, indicating a vegetated land cover adjacent to artificial structures or trees. Cluster 4 showed an NDVI of 0.72 and an SVF of 0.31, classified as an area mostly consisting of grassland and located very close to artificial structures or trees, with almost no visible perforations. As a result of cluster analysis based on NDVI and SVF into four cluster types, the land cover characteristics of vegetation and non-vegetation among the physical type characteristics were classified into NDVI characteristics of clusters 1, 3, 4 and cluster 2, and characteristics such as the degree of shielding according to the proximity of trees and artificial structures were judged to appropriately reflect the characteristics of open areas to areas very close to artificial structures and trees through the tendency of SVF values ​​to decrease as one moves from cluster 1 to 4.

[0081] FIG. 10 shows T within a park, a target site for NVI-met utilization according to an embodiment of the present invention. MRT This is a drawing showing the modeling results.

[0082] Figure 10 shows the TMRT modeling results for a height of 1.4m within a park at the ENVI-met utilization site according to an embodiment of the present invention. At 10:00, most areas of the site were found to have a temperature range of 65–70℃, while areas with dense trees were found to be 45℃ or lower, which is approximately 20℃ lower than the surrounding space. At 12:00, similar to the 10:00 modeling results, most areas of the site were analyzed to be in the 65–70℃ range, and areas with dense trees were predicted to be 50–55℃, which is approximately 10℃ higher than at 10:00. At 14:00, the overall average radiant temperature of the site was predicted to be very high, exceeding 70℃, and areas with dense trees were also confirmed to be slightly higher compared to the morning hours, ranging from 55–60℃. Finally, at 16:00, the average radiant temperature of most of the site was predicted to be similar to that of 14:00, exceeding 70℃, while artificial structures and areas with dense trees were found to be slightly lower, ranging from 50–55℃. The results of the thermal comfort analysis using ENVI-met modeling for the period of 10:00–16:00 revealed that the TMRT in high-density tree areas showed a difference of more than 30℃. This indicates that the urban forest canopy has a daily average of 5℃, and depending on solar energy incidence and surface conditions, the temperature can reach up to 33℃. MRT Similar to the results of the prior art disclosed as capable of reduction, it is determined that densely wooded areas contribute to the reduction of average radiant temperature. In addition, as shown in the results of Table 2, the prediction of the average radiant temperature at 16:00 in the present invention showed that the difference between densely wooded areas and other areas was 19.6℃, which was the largest among afternoon hours.

[0083] Table 2

[0084]

[0085] This is based on the analysis of the cooling effect by urban forests in summer, and as previously analyzed in the conventional data where the temperature difference between the forest and the city was greatest at 17:00, it is judged that the heat environment reduction effect in densely wooded areas becomes more pronounced after 16:00, when the sun's altitude decreases after 14:00, when the sun's altitude is highest.

[0087] FIG. 11 shows a physical space type and T by time period according to an embodiment of the present invention. MRT This is a diagram showing the results of the characteristic analysis.

[0088] Physical space types and T by time period MRT The results of the characteristic analysis are as shown in Figure 11 and Table 3.

[0089] Table 3

[0090]

[0091] 10 o'clock is T MRT Analysis showed that Cluster 2 had the highest temperature, with Cluster 1 at 64.2℃, Cluster 2 at 65.3℃, Cluster 3 at 63.5℃, and Cluster 4 at 63.6℃. At 12 o'clock, the maximum was 66.8℃ (Cluster 2) and the minimum was 65.8℃ (Clusters 1 and 3), and Cluster 2 type T MRT It was confirmed that was the highest. Next, for 14 and 16 o'clock, T by type MRT It was confirmed that the average value was over 70℃, which is about 5℃ higher than the morning time.

[0092] Meanwhile, consistent with the analysis results for the morning hours, Cluster Type 2 was found to have the highest average radiant temperature at 72.2°C at 14:00 and 72.3°C at 16:00. It is determined that the average radiant temperature increased as the incoming solar radiation energy increased with a higher SVF among physical spatial characteristics. When compared to Cluster Type 1, which has an SVF of 0.7 or higher similar to Cluster Type 2, it is judged that under conditions of high SVF, the average radiant temperature tends to decrease when NDVI increases.

[0094] FIG. 12 shows T over time according to physical spatial factors NDVI and SVF according to an embodiment of the present invention. MRT This is a diagram showing the results of the distribution analysis.

[0095] Figure 12 shows T over time according to physical spatial factors NDVI and SVF. MRT This is the result of the distribution analysis. It was confirmed that both NDVI and SVF factors showed a tendency for density to increase at 10:00 (63℃), 12:00 (63℃), 14:00 (67℃), and 16:00 (70℃). Furthermore, the NDVI and SVF values ​​showing a tendency to be dense were all found to be high, exceeding 0.6, confirming that types such as Cluster 1, which consists of open areas and vegetation, exhibit a similar trend to the thermal comfort change pattern over time. Therefore, it is judged that analysis of Cluster 1 type is possible when determining the thermal comfort distribution within the target site using UAVs.

[0097] FIG. 13 is a diagram showing the location of a point of low density in kernel density according to one embodiment of the present invention.

[0098] The location of the point of low density in kernel density is shown in Fig. 13. In a space where trees are mostly dense, T MRTIt showed a tendency for low correlation with. This is believed to be due to the limitations of UAV imaging, where it is difficult to acquire data for points obscured by water canopies or similar objects located at a height of 1.4m above the ground, as the present invention utilized NDVI and SVF based on UAV images captured at a 90° angle. Therefore, it is judged that this is caused by the difficulty in acquiring data for points obscured by water canopies or similar objects located at a height of 1.4m above the ground when compared with ENVI-met results modeled at a height of 1.4m above the ground. Accordingly, it is expected that in the future, it will be possible to identify the thermal comfort distribution for Cluster 3 and 4 types by collecting 3D physical environment data, such as ground LiDAR and SVF data from the surface using fisheye lenses, for points where UAV image acquisition is difficult, and correcting the accuracy of physical types based on UAV images. Furthermore, by collecting images under various conditions such as altitude, overlap, and shooting angle during UAV imaging, it will be possible to derive optimal conditions for classifying physical environment types in urban areas.

[0099] As a result of the thermal comfort distribution analysis according to an embodiment of the present invention, the possibility of identifying thermal comfort distribution based on UAV imagery for spatial types consisting of open areas and vegetation within urban areas was confirmed, and it was found that areas adjacent to trees are thermally more comfortable than open spaces. Based on this, it is determined that the use of artificial covering materials (marble, granite, wooden decks, etc.) for landscape purposes during the urban planning stage of creating plazas, shelters, and parks may act as a disadvantage in terms of human thermal comfort. Therefore, it is deemed necessary to create open spaces by considering natural covering materials such as grass and trees during the urban planning stage, and when using artificial covering materials, spatial planning should be carried out considering factors such as proximity to trees and buildings. Furthermore, the development of technology capable of rapidly identifying thermal comfort considering the physical environment near the surface of urban areas through UAV imagery is considered to be applicable as important basic data for identifying practical and tangible urban climate phenomena and for disaster and safety management.

[0100] The results of classifying physical space types based on UAV imagery for parks within urban areas according to an embodiment of the present invention and analyzing thermal comfort characteristics according to physical space types through comparison with ENVI-met thermal comfort results are summarized as follows.

[0101] The physical space types according to the embodiments of the present invention were classified into a total of four types, and physical space type characteristics appeared depending on the surface land cover material and the degree of proximity of trees and artificial structures. The ENVI-met thermal comfort results showed that T in high-density tree areas MRT It was analyzed that it has a difference of more than 30℃, and T at 16:00 MRT The prediction results showed that the difference between densely treeed areas and other areas was 19.6℃, the largest during the afternoon hours. Therefore, it is determined that the thermal environment reduction effect in densely treeed areas is relatively higher after 16:00 compared to daytime hours, and as a result of verifying the daytime solar radiation blocking effect by high-density urban forests, the tree density type [regarding] thermal comfort (T MRT It is judged that it contributes to the reduction. Finally, as a result of analyzing the characteristics of thermal comfort by physical space type and time of day, it was confirmed that for space types with high NDVI and high SVF, the use of UAVs showed a trend similar to the change pattern of thermal comfort by time of day. However, for areas densely populated with trees and artificial structures, the correlation with the UAV-based physical environment type was found to be relatively low.

[0102] In conclusion, according to an embodiment of the present invention, it was found that when classifying physical environment types using UAV image-based NDVI and SVF, it is possible to identify the distribution of thermal comfort in open areas with high NDVI and SVF, and it was confirmed that areas with dense trees are effective in reducing thermal comfort. Therefore, when planning cities with thermal comfort, it is judged that prioritizing landscape aspects and constructing open spaces such as plazas and parks with high population flow using artificial covering materials such as concrete and marble can act as a disadvantage in terms of thermal comfort for humans utilizing the space. Accordingly, it is judged that it is necessary to consider methods of constructing with natural covering materials such as grass and trees, and if artificial covering materials are used, spatial planning should be carried out considering SVF (proximity to trees and buildings). Furthermore, by referring to the results of studies conducted for the purpose of quantifying the thermal environment reduction effect by forests surrounding cities during the summer, it is possible to develop a more tangible urban heat island reduction planning technique and utilize it as important basic data for identifying practical and tangible urban climate phenomena and for disaster and safety management.

[0103] Accordingly, the present invention classified physical space types based on UAV imagery for parks within urban areas and analyzed thermal comfort characteristics according to physical space types through comparison with ENVI-met thermal comfort results.

[0104] According to an embodiment of the present invention, physical space types were classified into four types based on UAV-based NDVI (Normalized Different Vegetation Index) and SVF (Sky View Factor) characteristics. The ENVI-met thermal comfort results show the thermal comfort (T of a densely wooded area) MRT ) showed a difference of up to 30℃ or more compared to other areas, and during the afternoon hours, T in densely wooded areas and other areasMRT The difference was found to be greatest at 16:00, at 19.6℃. The analysis of UAV-based physical space types and thermal comfort characteristics by time of day confirmed that space types with high NDVI and SVF showed a trend similar to the change patterns of thermal comfort by time of day when using UAVs; however, areas densely populated with trees and artificial structures showed a relatively low correlation with UAV-based physical environment types. In conclusion, the feasibility of identifying thermal comfort distribution based on UAV imagery for space types consisting of open areas and vegetation was confirmed, and it was found that areas adjacent to trees were thermally more comfortable than open areas.

[0105] Therefore, during the urban planning stage, open spaces need to be created considering natural covering materials such as grass and trees, and when using artificial covering materials, spatial planning should be carried out considering factors such as proximity to trees and buildings. In the future, it is expected that rapid and accurate identification of urban climate phenomena and the establishment of urban plans considering thermal comfort will be possible through ground LiDAR and field measurement-based UAV image correction.

[0107] The device described above may be implemented as a hardware component, a software component, and / or a combination of a hardware component and a software component. For example, the device and components described in the embodiments may be implemented using one or more general-purpose or special-purpose computers, such as, for example, a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable array (FPA), a programmable logic unit (PLU), a microprocessor, or any other device capable of executing and responding to instructions. The processing unit may execute an operating system (OS) and one or more software applications executed on said operating system. Additionally, the processing unit may access, store, manipulate, process, and generate data in response to the execution of the software. For ease of understanding, the processing unit may be described as being used as a single unit, but those skilled in the art will understand that the processing unit may include multiple processing elements and / or multiple types of processing elements. For example, the processing unit may include multiple processors or one processor and one controller. In addition, other processing configurations, such as parallel processors, are also possible.

[0108] Software may include computer programs, code, instructions, or a combination of one or more of these, and may configure a processing unit to operate as desired or instruct the processing unit independently or collectively. Software and / or data may be embodied in any type of machine, component, physical device, virtual equipment, computer storage medium, or device so as to be interpreted by the processing unit or to provide instructions or data to the processing unit. Software may be distributed over networked computer systems and may be stored or executed in a distributed manner. Software and data may be stored on one or more computer-readable recording media.

[0109] The method according to the embodiment may be implemented in the form of program instructions that can be executed through various computer means and recorded on a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, etc., either alone or in combination. The program instructions recorded on the medium may be those specifically designed and configured for the embodiment, or they may be those known and available to those skilled in the art of computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tapes; optical recording media such as CD-ROMs and DVDs; magneto-optical media such as floptical disks; and hardware devices specifically configured to store and execute program instructions, such as ROM, RAM, and flash memory. Examples of program instructions include machine code, such as that generated by a compiler, as well as high-level language code that can be executed by a computer using an interpreter, etc.

[0110] Although the embodiments have been described above with reference to limited examples and drawings, those skilled in the art can make various modifications and variations from the description above. For example, suitable results can be achieved even if the described techniques are performed in a different order than described, and / or the components of the described system, structure, device, circuit, etc. are combined or assembled in a form different from described, or replaced or substituted by other components or equivalents.

[0111] Therefore, other implementations, other embodiments, and equivalents to the claims also fall within the scope of the claims set forth below.

Claims

Claim 1 The method comprises: a step of classifying physical environment types for a target site for thermal environment characteristic analysis according to unmanned aerial vehicle-based NDVI (Normalized Different Vegetation Index) and SVF (Sky View Factor) characteristics; a step of analyzing thermal comfort characteristics of the target site using microclimate modeling (ENVI-met); and a step of analyzing the distribution of unmanned aerial vehicle-based thermal comfort by comparing the classified physical environment types and the analyzed thermal comfort characteristics, wherein the step of analyzing thermal comfort characteristics of the target site using microclimate modeling (ENVI-met) includes thermal comfort (T) for measuring and expressing radiant energy affecting the thermal sensation of the human body among the simulation results using microclimate modeling (ENVI-met). MRT The step of extracting ) data, converting it into raster data to analyze the thermal comfort characteristics of the target site, and analyzing the unmanned aerial vehicle-based thermal comfort distribution by comparing the classified physical environment types and the analyzed thermal comfort characteristics involves the thermal comfort (T) extracted from a simulation using microclimate modeling (ENVI-met). MRT Using ), a distribution map of the error between the modeled values ​​and the predicted values ​​is constructed, and thermal comfort (T) by physical environment type and time period is MRT ), hourly thermal comfort (T MRT Analyzing the distribution, if the subject site corresponds to a first physical environment type among the classified physical environments where NDVI and SVF are higher than predetermined criteria, the time-varying thermal comfort (T) extracted based on the characteristics of the first physical environment type and a simulation utilizing the microclimate modeling (ENVI-met) MRT While determining that the correlation between the change patterns of ) is relatively higher, if the subject site corresponds to the second physical environment type among the classified physical environments, which is an area densely populated with trees and artificial structures, the time-dependent thermal comfort (T) extracted based on the characteristics of the second physical environment type and a simulation utilizing the microclimate modeling (ENVI-met) MRT A thermal environment characteristic analysis method that determines the correlation between the change patterns of ) as relatively lower. Claim 2 In claim 1, the step of classifying physical environment types according to unmanned aerial vehicle-based NDVI (Normalized Different Vegetation Index) and SVF (Sky View Factor) characteristics for a target site for thermal environment characteristic analysis is a thermal environment characteristic analysis method that collects unmanned aerial vehicle images by capturing multispectral images and optical images through an unmanned aerial vehicle to classify physical environment types. Claim 3 A method for analyzing thermal environment characteristics according to paragraph 2, wherein the RGB and multispectral images of the collected unmanned aerial vehicle images are produced as orthophotos using software, and a Digital Surface Model (DSM) is produced using optical images for SVF analysis. Claim 4 In paragraph 3, the step of classifying the physical environment type according to the unmanned aerial vehicle-based NDVI (Normalized Different Vegetation Index) and SVF (Sky View Factor) characteristics for the target site for thermal environment characteristic analysis comprises calculating the NDVI and SVF characteristics based on the produced unmanned aerial vehicle imagery, and utilizing land use and cover types and the degree of density of buildings and trees according to the NDVI and SVF characteristics as classification factors for the physical environment type of the target site. Claim 5 A method for analyzing thermal environment characteristics according to claim 4, wherein the above NDVI is a unitless radiation value calculated through reflection values ​​in the near-infrared and red band regions, the above SVF requires urban morphology data as an indicator representing the degree of building density by analyzing the ratio of the area where the sky is visible at a surface observation point, and is calculated through a terrain analysis function utilizing a DSM based on high-resolution RGB images of an unmanned aerial vehicle, and for the classification of physical environment types, the number of physical environment type clusters is derived through Python-based inertia analysis, and then K-means clustering is performed to classify physical environment types. Claim 6 delete Claim 7 In claim 1, the step of analyzing the thermal comfort characteristics of the target site using the microclimate modeling (ENVI-met) comprises a method for analyzing thermal environment characteristics in which an input file (.IN) for physical environmental factors in the microclimate modeling simulation is constructed by inputting building shape, height, vegetation height and species, and covering material through RGB optical images of an unmanned aerial vehicle and field surveys, and a simulation condition file (.CF) is constructed by setting coordinates, grid size, initial wind speed and direction, and temperature. Claim 8 delete Claim 9 It comprises: a physical environment type classification unit that classifies physical environment types according to unmanned aerial vehicle-based NDVI (Normalized Different Vegetation Index) and SVF (Sky View Factor) characteristics for a target site for thermal environment characteristic analysis; a thermal comfort characteristic analysis unit that analyzes thermal comfort characteristics of the target site using microclimate modeling (ENVI-met); and a thermal comfort distribution analysis unit that analyzes the unmanned aerial vehicle-based thermal comfort distribution by comparing the classified physical environment types and the analyzed thermal comfort characteristics, wherein the thermal comfort characteristic analysis unit includes thermal comfort (T) for measuring and expressing radiant energy affecting the thermal sense of the human body among the simulation results using microclimate modeling (ENVI-met). MRT ) extracts data, converts it into raster data to analyze the thermal comfort characteristics of the target site, and the thermal comfort distribution analysis unit analyzes the thermal comfort (T) extracted from a simulation using microclimate modeling (ENVI-met). MRT Using ), a distribution map of the error between the modeled values ​​and the predicted values ​​is constructed, and thermal comfort (T) by physical environment type and time period is MRT ), hourly thermal comfort (T MRT Analyzing the distribution, if the subject site corresponds to a first physical environment type among the classified physical environments where NDVI and SVF are higher than predetermined criteria, the time-varying thermal comfort (T) extracted based on the characteristics of the first physical environment type and a simulation utilizing the microclimate modeling (ENVI-met) MRT While determining that the correlation between the change patterns of ) is relatively higher, if the subject site corresponds to the second physical environment type among the classified physical environments, which is an area densely populated with trees and artificial structures, the time-dependent thermal comfort (T) extracted based on the characteristics of the second physical environment type and a simulation utilizing the microclimate modeling (ENVI-met) MRT A thermal environment characteristic analysis device that determines the correlation between the change patterns of ) as relatively lower. Claim 10 In claim 9, the physical environment type classification unit is a thermal environment characteristic analysis device that collects unmanned aerial vehicle images by capturing multispectral images and optical images through an unmanned aerial vehicle to classify physical environment types. Claim 11 In claim 10, the physical environment type classification unit is a thermal environment characteristic analysis device that produces an orthophoto using S / W from the RGB and multispectral images of the collected unmanned aerial vehicle images, and produces a Digital Surface Model (DSM) using optical images for SVF analysis. Claim 12 In claim 11, the physical environment type classification unit calculates NDVI and SVF characteristics based on the produced unmanned aerial vehicle image, and the thermal environment characteristic analysis device utilizes land use and cover type and the degree of density of buildings and trees according to the NDVI and SVF characteristics as physical environment type classification factors of the target site. Claim 13 A thermal environment characteristic analysis device according to claim 12, wherein the above NDVI is a unitless radiation value calculated through reflection values ​​in the near-infrared and red band regions, the above SVF requires urban morphology data as an indicator representing the degree of building density by analyzing the ratio of the area where the sky is visible at a surface observation point, and is calculated through a terrain analysis function utilizing a DSM based on high-resolution RGB images of an unmanned aerial vehicle, and for physical environment type classification, the number of physical environment type clusters is derived through Python-based inertia analysis, and then K-means clustering is performed to classify physical environment types. Claim 14 delete Claim 15 In claim 9, the thermal comfort characteristic analysis unit is a thermal environment characteristic analysis device in which the input file (.IN) for physical environmental factors in the microclimate modeling simulation is constructed by inputting building shape, height, vegetation height and species, and covering material through RGB optical images of an unmanned aerial vehicle and field surveys, and the simulation condition file (.CF) is constructed by setting coordinates, grid size, initial wind speed and direction, and temperature. Claim 16 delete

Images