Digital evaluation and analysis method for urban landscape overlook system based on color coding

Abstract

The invention discloses a digital evaluation and analysis method for an urban landscape overlook system based on color coding, and the method comprises the steps: collecting map data of a research region, and making a basic base map; determining the form of the research range and an urban landscape overlook system; constructing a background environment of the overlook object by constructing a geographical environment, a background building environment and other required visual elements; performing color coding on the overlook object and the overlook system background environment to obtain a city three-dimensional information model; performing automatic street view data simulation acquisition in an urban simulation environment according to the sampling viewpoint to obtain a scene picture; performing image processing on the scene picture, and performing interpretation according to picture color coding information; performing data statistics and planning design application; the method solves the problems that quantitative evaluation cannot be carried out from the perspective of people, and the evaluation accuracy is affected by large limiting factors of analysis.

Description

Technical Field

[0001] The present invention relates to the field of visual analysis technology, and in particular to a digital evaluation and analysis method for a city landscape viewing system based on color coding. Background Art

[0002] There have been studies on the visual analysis of urban landscape viewing systems at home and abroad. The main quantitative evaluation methods are: one is landscape visual calculation and visibility analysis based on GIS platform, and the other is quantitative calculation of visual elements through data analysis of Internet street view images. The main defects of the above two methods are as follows: 1. GIS analysis focuses on the visibility of "objects" and ignores the information perception of "people". Spatial perception is the process of people's comprehensive perception of the concrete content of visual objects such as form, color, and environment. The basic principle of GIS analysis method is to abstract people's viewpoints into single points (static viewpoints) or multiple points (moving viewpoints), and then analyze and judge whether the surrounding space is visible with the point as the center, and then accumulate the number of times the space is visible from all viewpoints to achieve quantitative evaluation. This method can evaluate the visibility of an area or building, but it cannot quantitatively describe the visual information at the viewpoint, that is, it cannot give a quantitative evaluation of the quality of the visual landscape from a human perspective.

[0003] 2. Data sources are limited, and the breadth and accuracy of analysis are greatly limited. Another way of landscape visual analysis is to use image recognition to analyze street view photos. Compared with GIS analysis, which can reflect the concrete information of visual objects, it is subject to the location and quality of data collection, and the analysis is greatly limited. The existence of Internet street view photo data depends on the location of the data collected by the Internet company, and the data quality (the clarity of the photo and whether it is blocked by irrelevant elements) directly affects the accuracy of the image recognition results. Image recognition technology also requires a large number of physical photos of the viewing object to be collected and trained to ensure that the object can be effectively identified in the street view photos. Summary of the invention

[0004] 1. Technical issues to be resolved Based on the above problems, the present invention provides a digital evaluation and analysis method for a city landscape viewing system based on color coding, which solves the problem that the evaluation accuracy is greatly affected by limiting factors that cannot be quantitatively evaluated and analyzed from a human perspective.

[0005] (II) Technical solution Based on the above technical problems, the present invention provides a digital evaluation and analysis method of a city landscape viewing system based on color coding, comprising: S1. Collect map data of the study area and make a basic base map; S2. Determine the form of the research scope and the urban landscape viewing system. The urban landscape viewing system includes sampling viewpoints and viewing objects; S3. Construct the background environment of the viewing system by constructing the geographical environment, background architectural environment and other required visual elements; S4, color-coding the viewing object and the viewing system background environment respectively to obtain a three-dimensional model with color visual information; S5, performing automated street view data simulation collection in an urban simulation environment according to the sampling viewpoint to obtain a scene photo, wherein the urban simulation environment includes a viewing object and a background environment of the viewing object; S6, performing image processing on the scene photo and interpreting it according to the color coding information of the picture; S7. Data statistics and planning design applications.

[0006] (III) Beneficial effects The above technical solution of the present invention has the following advantages: (1) The present invention focuses the research perspective of the landscape system on people, making up for the technical defects of traditional GIS visual analysis that is centered on "objects". GIS technology is used to build a simplified urban white model environment of the real space, and the color coding method is introduced to use the CGA rule modeling method to assign different color visual information to landscape objects according to unique ID values, making up for the problem that traditional street view images have too much visual interference and insufficient recognition effect in terms of information breadth and accuracy. Finally, relying on the viewpoint, automatic collection is carried out in the simulated environment, and image processing technology is used to accurately identify the color with coded information to transmit visual information, providing a basis for the quantitative evaluation of the urban landscape system and assisting planning and design; (2) The present invention can autonomously collect and generate data for analysis, expand spatial freedom, reduce the limitations of data collection location and quality on analysis, and improve the accuracy of evaluation; at the same time, it can simplify unnecessary visual elements through the model and increase the accuracy of image processing; (3) The present invention avoids the complex process of understanding mathematical principles. The principle is simple and consistent with the human visual cognitive process. At the same time, it can increase the features of the viewed object in the scene according to the evaluation requirements and improve the amount of visual information. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] The features and advantages of the present invention will be more clearly understood by referring to the accompanying drawings, which are schematic and should not be construed as limiting the present invention in any way. In the accompanying drawings: Figure 1 It is a flow chart of a digital evaluation and analysis method of a city landscape viewing system based on color coding according to an embodiment of the present invention; Figure 2 is a detailed flow chart of step S2 of an embodiment of the present invention; Figure 3 is a detailed flow chart of steps S3-S4 of an embodiment of the present invention; Figure 4 is a detailed flow chart of steps S5-S6 of an embodiment of the present invention; Figure 5 This is a schematic diagram of planar range viewpoint sampling according to an embodiment of the present invention; Figure 6 This is a schematic diagram of linear range viewpoint sampling according to an embodiment of the present invention; Figure 7 It is a script of the function mkColor of the embodiment of the present invention; Figure 8 A schematic diagram for generating an information model according to an embodiment of the present invention; Fig. 9 A script of the three-dimensional information model function mkExtrusion of an embodiment of the present invention; Fig.10 A script for the color analysis function colorhsv of an embodiment of the present invention; Fig.11 The figure is a schematic diagram of visual expression of regional landscape visibility assessment according to an embodiment of the present invention. DETAILED DESCRIPTION

[0008] The specific implementation of the present invention is further described in detail below in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0009] The embodiment of the present invention is a digital evaluation and analysis method of a city landscape viewing system based on color coding, such as Figure 1-4 As shown, the following steps are included: S1. Collect map data of the study area and make a basic base map; For the study area, data is collected and a data base map is produced through Internet maps, oblique photography, etc. The map data involved are mainly Internet map image data, building outlines, road alignments, DEM terrain, POI, AOI and other open source data, as well as street view photos or independently collected landmark building photos as information reference supplements. Among them, most of the building outline data can be downloaded from open source maps (such as OpenStreetMap). The downloaded building outline data is accompanied by a height attribute field. Some missing buildings can be obtained by manual drawing, and the height attribute is determined based on the building shadow length of the map image or the street view photo.

[0010] S2. Determine the form of the research scope and the urban landscape viewing system. The urban landscape viewing system includes sampling viewpoints and viewing objects: S21. Form of determining the scope of research; According to the needs of assessment and planning, the scope of research can be in various forms, including surface, line or point scope. Among them, the surface scope is to define a certain range for the region to simulate the situation where people move freely in the regional space to look at the surrounding environment; the linear scope refers to the situation along the linear space, such as along the road and river bank, to simulate the situation where people move along the linear space to look at the surrounding environment; the point scope is to simulate the situation where people look at the surrounding environment at a specific location, such as the viewing point of the landscape. The above scopes create surface, line and point layers in ArcGIS software and add corresponding elements.

[0011] S22. Determine the sampling viewpoints of the urban landscape viewing system according to the form of the research scope: The sampling viewpoints are determined according to the research scope, that is, the different forms of surface, line or point scope, as follows: Among them, the area range viewpoints are generated by the ArcGIS Create Fishnet Point tool to generate several viewpoints, and the grid points covered by buildings are deleted, and then the points are supplemented as needed to form sampling viewpoints, such as Figure 5 As shown; The linear range viewpoints are obtained by performing the following operations on the linear features in ArcGIS software: 1) performing the "Feature to Point" operation in sequence to obtain the midpoint of the line feature, and then performing the "Split Line at Feature Turn Point" operation to achieve an even division of the original line segment by length; 2) repeating the operation 1 on the segmented line segment, and determining the number of repetitions N according to the density of the required sampling viewpoints to obtain 2N equally divided and equally spaced sampling viewpoints, such as Figure 6 As shown; The point range can be directly added by drawing the determined sampling viewpoints as point features.

[0012] The above-mentioned various sampling viewpoints are uniquely numbered by adding the attribute field ID for easy recording.

[0013] S23. Extracting viewing objects of the urban landscape viewing system: The viewing object refers to the landscape elements in the urban viewing system, including artificial landscape elements (such as landmark buildings) and natural landscape elements (such as mountain landscape). The above landscape elements are determined as follows: S231. Extract natural landscape elements, where the natural landscape elements include mountain landscape.

[0014] Determine the geographical location of the landscape element based on the POI and AOI of the landscape type, and extract the natural landscape element data in the following ways: S2311, converting terrain raster data into vector line data with elevation information, and then generating closed contour surface feature data through the feature-to-surface tool; In ArcGIS software, the terrain raster data is converted into vector line data with elevation information through the 3D Analyst tool\Raster Surface\Contour tool, and then the closed contour surface feature data is generated through the Feature to Surface tool; S2312. Delineate the natural landscape element range surface based on the landscape AOI range information.

[0015] Based on the landscape AOI range information, the range of natural landscape elements (surface elements) is delineated. For some natural mountains without AOI data, contour lines can be generated from terrain raster data. Closed contour lines are selected in the area where the landscape POI point is located to generate surface elements. S2313, clipping the contour lines (area elements) within the range of the natural landscape elements as natural landscape element data; S2314. Natural landscape elements and terrain are segmented, and the terrain raster data is integrated and updated.

[0016] After extracting natural landscape elements (mountain elements), in order to avoid overlap with the terrain raster data within the range after superposition, the original terrain raster data needs to be updated. The specific method is: The contour values ​​within the range of the mountain landscape element are traversed in turn, the minimum value is screened out, the range of the mountain landscape element is converted into raster data according to the minimum elevation value, and then merged with the terrain raster data through the mosaic tool, so as to achieve the effect of dividing the mountain landscape element and the terrain.

[0017] The above realizes the extraction of natural landscape elements and generates terrain raster data for separating natural landscape elements. Among them, AOI (Area of ​​Interest) refers to the interest surface in the Internet electronic map, which is mainly used to express regional geographical entities in the map, such as a residential area, a university, an office building, an industrial park, a comprehensive shopping mall, a hospital, a scenic spot or a gymnasium, etc.; POI (Point of Interest), on the electronic map, POI represents restaurants, supermarkets, government agencies, tourist attractions, transportation facilities, etc.

[0018] S232: extracting artificial landscape elements, wherein the artificial landscape elements include landmark buildings.

[0019] The general geometric characteristics of landmark buildings are high building height, large building land area, and generally commercial office, public service, cultural heritage and other purposes, and they are well-known in the local area. Based on the above characteristics, the following methods can be used to determine landmark buildings: S2321. Filter building outlines according to their geometric properties; The data is obtained by sorting the building height (BuildingHeight) and area (Shape_Area) fields of the building outline data from high to low and then taking the top 10% of the data.

[0020] S2322. Use reverse geocoding to obtain building functions and filter out the outlines of buildings used as commercial offices and public buildings; The longitude and latitude (WGS84 coordinates) of the center point of the building outline are used as request parameters. The coordinates are first converted into coordinate values ​​under the Internet map coordinate system through the Internet map coordinate system conversion API. The coordinate values ​​are then used to obtain the building information through the Internet map reverse geographic coordinates, including the building's purpose, address, building name, etc., and recorded by establishing corresponding data attribute fields; the relevant outline data for commercial offices and public buildings are screened out.

[0021] S2323. Use POI, AOI and collected information, supplemented by tourism software APP to select the outlines of historical and cultural landmark buildings; Through POI, AOI and collected information, supplemented by travel software APP, the outlines of historical and cultural landmark buildings can be found, and their attribute information such as building purpose, address, and building name can be recorded.

[0022] S2324. After verifying the popularity of the selected building outlines, they will be used as landmark building elements; According to the landmark buildings and related information selected above, their local popularity is verified through surveys, visits, the Internet and other means.

[0023] S3. Construct the background environment of the viewing object by constructing the geographical environment, background architectural environment and other required visual elements: S31. Construct geographic environment, including terrain raster data, road element data and block data a) loading the terrain raster data of the basic base map; b) Loading roads: Use the buffer tool to perform buffer operations on road line features according to the width value determined by the road level to generate road surface feature data; c) Loading block data: Use the Feature to Polygon tool of ArcGIS software to generate block faces from the enclosed area of ​​the closed road line features, establish a block ID attribute field to uniquely identify each block face; assign attribute information of the block ID where the building outline data is located according to the spatial position.

[0024] S32. Generate background building environment: After erasing the outline of the landmark building from the original building outline data, the outline data of the background building is used as the outline data. The building outline vector data is imported into the scene, and the CityEngine software uses the Extrude function in the CGA rule to automatically stretch the building block according to the attribute value of the height attribute in the building outline vector data to generate the background building environment.

[0025] S33. Import viewing object: 1) Encoding the mountain landscape data (natural landscape elements) and landmark building data (artificial landscape elements). The mountain landscape data and landmark building data with attribute information generated in step S2 are uniformly encoded.

[0026] 2) Import the CityEngine software integration data. The above landscape element data will be modeled in the color coding method of step S4 in the following.

[0027] S34. Create other required visual elements: Other required visual elements refer to model elements that actually exist in the scene and affect the evaluation results when simulating people's viewing behavior, such as roadside trees with occlusion in the visual evaluation scene along the road, densely populated woods in the park landscape evaluation, etc. These elements can be located first according to the evaluation needs, and then imported into the scene of CityEngine software as point element vector data, and added by inserting the corresponding white model at the point position by writing CGA rule files.

[0028] S35, loading terrain grid data according to terrain layer, loading the road surface element data and block surface element data, and loading other required visual elements and viewing objects of the background building environment in sequence, and then vertically projecting them on the terrain grid graphics of the geographical environment through the AlignShapes to Terrain function so as to make them close to the terrain, thereby constructing the viewing object background environment.

[0029] S4. Color-code the viewing object and the viewing system background environment to obtain a three-dimensional model with color visual information: Color coding is a type of coding that uses color as a code for visual information. In order to obtain information from the image generated by the sampling viewpoint, the attribute information of the viewing object needs to be encoded into specific colors according to certain rules. The corresponding color coding parsing process is to reversely interpret the color information into attribute information.

[0030] The specific implementation process of color coding in the present invention is: in the process of generating the three-dimensional information model of landscape elements, the information attribute value (or category code value) is read and converted into a unique corresponding RGB color value through calculation, and then the color is dyed to the specified model area through the CGA rule built-in function color, so that the model obtains the color visual information. This color coding adopts the HSV color mode to uniquely represent the specific information value with H (hue), V (saturation), and S (brightness). The main operation process is as follows: S41. Write a color coding function through CGA rules and generate a three-dimensional model of the landscape element, where the landscape element is the viewing object.

[0031] S411, setting attribute parameters. Set the attribute parameters of the current element in the CGA rule file, such as the object ID number, the expression such as attr id = 0, where the ID number attribute variable name is consistent with the attribute name of the landscape element data. In the CityEngine software, the specific value of the variable defaults to 0, and once the function is applied, it automatically matches the attribute field with the same attribute name and assigns the ID value of the data to id; another variable is the maximum ID value, the expression such as attr idMax = 100, its value is the maximum ID value of all landscape elements.

[0032] S412: Convert the ID value into an HVS color value.

[0033] In the HSV color mode, the value range of H (hue) in the CGA rule of CityEngine software is 0-360. By dividing 0-360 into idMax+1 parts and taking the middle value of the corresponding id interval, the ID variable value is converted into an interval number of 0-360 to establish a corresponding relationship with the H value. The other V (saturation) and S (brightness) are fixed to 1. The calculation formula of the H value is as follows: 180 / (idMax+1) + (id-1)*360 / (idMax+1). This calculation method obtains the corresponding color code of the current feature ID, which is not easy to conflict with the background color value (white or gray, where gray is the deviation of white caused by shadow) and the identification color (black) under the premise of ensuring the maximum distinction.

[0034] S413: Color the viewing object according to the HVS color value.

[0035] The HVS color value in S412 is used as the parameter H input value of the built-in function colorHSVToHex of the CGA rule, and the other V and S inputs are 1, and the output result is the hexadecimal color value of the color value. The three functions colorHexToR, colorHexToG, and colorHexToB are used to obtain the Red, Green, and Blue channel values ​​of the hexadecimal color in the RGB color mode respectively. Therefore, the color function can be called to color the model area, realizing the function function of converting the information attribute variable to the single color visual information. The function mkColor is shown as an example. Figure 7 As shown; S414, constructing an information model generation function. Landscape elements generate a three-dimensional model by establishing two layers of models, the main layer and the identification layer. The main layer is the main part of the model of the landscape element. It is a three-dimensional model body that is stretched to the building height by the stretching function extrude and the color coding information is assigned to the corresponding color by the element ID attribute. The identification layer is based on the main layer and stretched by 2 unit heights (for example, the variable name is set to unitHeight). The unit height is set according to the normal display in the graphic as the standard, which can be used as the conversion unit between pixels and actual size.

[0036] The two layers of the logo layer are respectively assigned black (logo color) and white (background color). The upper white layer is set to further prevent the adhesion of similar colors and to separate overlapping objects; the lower black layer is used to quickly mark the position of the object with a uniform color. At the same time, since the actual simulated height of this layer is known, the actual height of the main layer can be deduced through the pixel height of the image, which serves as a scale.

[0037] Assuming that the maximum ID is 10, the information model generation diagram is as follows: Figure 8 As shown; Combined with the color coding function of the figure, the three-dimensional information model function mkExtrusion of the above landscape elements is expressed as follows Fig. 9 As shown; S42, the background model sets the background color to simplify the scene. The three-dimensional model of the landscape elements is set in color according to the coding mode, and the three-dimensional model of other building outlines is set to (1,1,1) or white using the color function, so as to obtain a scene that highlights the visual object and weakens the background information. Due to the occlusion of the scene shadow, some areas covered by the shadow appear gray, so they are also set as the background model color.

[0038] S5. Automated street view data simulation collection is performed in an urban simulation environment according to the sampling viewpoint to obtain scene photos, wherein the urban simulation environment includes a viewing object and a viewing object background environment: Import the sampled viewpoints into the scene created in step S4 and write a Python automation script to implement the following operations: S51, picking up a sampling point in numerical order and moving the camera lens to the point; S52, setting the camera lens parameters according to the general observation characteristics of the average height of a person to an elevation angle of 30°, a horizontal field of view of 120°, and a height of 1.6 m. Other perspective roles may also be set as needed, such as a child's perspective.

[0039] S53. At the sampling point, the scene photos are captured respectively when the lens is rotated along the y-axis by 0°, 90°, 180°, and 270° (i.e. -90°), and are named in a unified letter identifier, viewpoint ID, and viewing angle content format in a unified “_” connection manner, such as IMG_01_90.png, and saved in a specified folder.

[0040] S6. Perform image processing on the scene photo, and reversely interpret the colors in the photo according to the color coding rules to obtain information represented by the colors: Perform image processing on the scene image generated in step S5, and use the OpenCV module to write a Python script to interpret the color coding information of the image.

[0041] S61, identifying a logo layer in the image.

[0042] 1) Use the color of the logo layer (black) as the ROI (region of interest), specify the corresponding interval of black in the HSV color mode, and select all logo layers through the OpenCV module function cv2.inRange. 2) Get the image position information of all ROIs: Use the cv2.threshold function to binarize the image and use cv2.findContours to detect the contour information of all ROIs. The detection information is stored in contours; traverse contours and call cv2.boundingRect one by one to obtain the coordinate value information of the circumscribed rectangle of the contour, including x, y, w, h, where x, y are the coordinates of the upper left corner, w, h are the width and height of the matrix, so as to determine the position (x, y) of the viewed object in the image and the height h of the model corresponding to the image.

[0043] S62, obtain a distinguishable binary background image. Take the background building and the environment color (white, gray) as ROI (region of interest), specify the corresponding interval of white and gray in the HSV color mode (the interval where the white and gray color space is basically continuous), select the ROI area through the OpenCV module function cv2.inRange, and perform binarization processing through the cv2.threshold function. After processing, the area with an image pixel value of 1 belongs to the original background, and the area with a value of 0 belongs to the imaging area of ​​the landscape element of the image.

[0044] S63. Calculate the main layer image area of ​​the element. Construct a convolution kernel of h*h with the circumscribed rectangle of the identification layer calculated in 1), and take the coordinates of the upper left corner of the element identification as the integer of w / 2 to the right, that is, the center of the first row of the identification layer ROI (x, y + int (w / 2)) as the starting point to scan the binary background image downward from h. Stop scanning only when the pixel values ​​in the convolution kernel are all 1, record the number of scans n, and calculate the position information of the main layer ROI area range as width w, height n*h, upper left corner coordinates as the position of the identification layer moved down one layer height, that is (x+h, y), and area n*w*h. The above image information can be used to obtain the actual visible range according to the actual size unitHeight corresponding to h.

[0045] S64. Calculate the color of the main layer of the feature and parse the ID value.

[0046] To ensure the accuracy of color analysis, the center area of ​​the main layer ROI in the original image is taken as the color value of the parsed feature ID value. The coordinates of the center point are (x +h + int(h*n / 2)), y+int(w / 2)), where the int function represents rounding to obtain the value in the HSV color mode at that point. In the Python programming environment, the value range of H is (0,180), so according to the function logic of the CGA rule during encoding, the entire interval is divided into idMax+1 equal parts. The color value is the median of the idth interval, where idMax is the maximum ID value of all elements set when the model is encoded using the CGA rule, and id is the ID code of the model. Therefore, by dividing the H value by the equally divided interval and then taking the value downward and adding 1, the index number of the interval, that is, the feature id value, can be obtained, and the feature id value can be parsed. Python script functions such as Fig.10 As shown, colorhsv is the hsv color value of the center position of the main layer ROI; According to the above operations, the following information can be obtained: 1) Image analysis information: the ID number of the landscape element presented in the image, the position, length, height and visible area of ​​the visible area of ​​the image; 2) Viewpoint information: by extracting the number information of the photo name and the viewpoint parameters of the CityEngine software Python script, the viewpoint ID, observation direction, elevation angle, horizontal field of view, visual height and other perspective simulation information of the generated image can be obtained, and the ID can be associated with the actual viewpoint position, elevation and other information; 3) Actual information: According to the conversion relationship between image pixels and actual size, the actual visible height, area, visible real height and all actual attribute information associated with the feature ID of the viewpoint can be calculated.

[0047] S65. The analysis contents of the above pictures are uniformly entered into text for analysis and statistics.

[0048] S7. Conduct statistical analysis on the interpreted information and apply it to planning and design.

[0049] (1) Regional landscape visibility assessment. Extract the viewpoint ID and the visible area of ​​the landscape element image from the text information to assess the regional landscape visibility. Summarize the visible area of ​​the element image by viewpoint ID, and use the attribute connection tool to associate the viewpoint with the statistical results under the viewpoint. Based on ArcGIS, the results are visually expressed through gridding, as shown in the following figure. Fig.11 As shown, the color depth indicates the visibility of the current landscape elements, providing a basis for the selection of viewing points in planning and design.

[0050] (2) Evaluation of the visual integrity of regional landscapes. The landscape element ID, landscape visible height and actual height are extracted from the text information to evaluate the visual integrity of the regional landscape. The visual integrity refers to the visible area ratio of the landscape elements that can be seen, that is, the visible height / actual height. The selected data are aggregated by landscape element ID. The aggregation method is to take the average of the visible area ratio of the elements, associate the viewpoint position by attribute through ArcGIS software, and realize the visual expression of the results through ArcGIS gridding. The evaluation results provide a reference for selecting important landscape elements and comparing the landscape effects under different height controls.

[0051] (3) Landscape element orientation analysis. Viewpoint IDs and landscape element IDs were extracted from text information, and OD analysis was established through ArcGIS to express the degree of visual connection between landscape elements and regional viewpoints. The evaluation results can provide a reference for setting the location of landscape view corridors based on landscape resources.

[0052] In summary, the digital evaluation and analysis method of the urban landscape viewing system based on color coding has the following beneficial effects: (1) The present invention focuses the research perspective of the landscape system on people, making up for the technical defects of traditional GIS visual analysis that is centered on "objects". GIS technology is used to build a simplified urban white model environment of the real space, and the color coding method is introduced to use the CGA rule modeling method to assign different color visual information to landscape objects according to unique ID values, making up for the problem that traditional street view images have too much visual interference and insufficient recognition effect in terms of information breadth and accuracy. Finally, relying on the viewpoint, automatic collection is carried out in the simulated environment, and image processing technology is used to accurately identify the color with coded information to transmit visual information, providing a basis for the quantitative evaluation of the urban landscape system and assisting planning and design; (2) The present invention can autonomously collect and generate data for analysis, expand spatial freedom, reduce the limitations of data collection location and quality on analysis, and improve the accuracy of evaluation; at the same time, it can simplify unnecessary visual elements through the model and increase the accuracy of image processing; (3) The present invention avoids the complex process of understanding mathematical principles. The principle is simple and consistent with the human visual cognitive process. At the same time, it can increase the features of the viewed object in the scene according to the evaluation requirements and improve the amount of visual information.

[0053] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the embodiments of the present invention are described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations shall fall within the scope defined by the appended claims.

Claims

1. A digital evaluation and analysis method for a color-coded urban landscape viewing system, characterized in that: include: S1. Collect map data of the study area and make a basic base map; S2. Determine the form of the research scope and the urban landscape viewing system. The urban landscape viewing system includes sampling viewpoints and viewing objects; S3. Constructing the background environment of the viewing object by constructing the geographical environment, the background architectural environment and other required visual elements; the other required visual elements are model elements that actually exist in the scene and affect the evaluation results when simulating the viewing behavior of a person; S4, color-coding the viewing object and the viewing system background environment respectively to obtain a three-dimensional model with color visual information; S5, performing automated street view data simulation collection in an urban simulation environment according to the sampling viewpoint to obtain a scene photo, wherein the urban simulation environment includes a viewing object and a background environment of the viewing object; S6, performing image processing on the scene photo, and reversely interpreting the colors in the photo according to the color coding rule to obtain information represented by the colors; S7. Conduct statistical analysis on the interpreted information and apply it to planning and design.

2. The digital evaluation and analysis method of the urban landscape viewing system based on color coding according to claim 1 is characterized in that: The map data includes Internet map image data, building outlines, road alignments, DEM terrain, POI, AOI open source data, street view photos, and independently collected landmark building photos.

3. The digital evaluation and analysis method of the urban landscape viewing system based on color coding according to claim 1 is characterized in that: The step S2 comprises: S21. Determine the form of the research scope, wherein the form of the research scope includes a planar scope, a linear scope, or a point scope; S22. Determine the sampling viewpoints of the urban landscape viewing system according to the form of the research scope; S23: extracting viewing objects of the urban landscape viewing system, wherein the viewing objects include natural landscape elements and artificial landscape elements.

4. The digital evaluation and analysis method of the urban landscape viewing system based on color coding according to claim 3 is characterized in that: Step S22 includes: The viewpoints of the planar range are generated by the ArcGIS Create Fishnet Point Tool, by deleting the grid points covered by buildings and then supplementing the points as needed to form sampling viewpoints; The viewpoints of the linear range are obtained by performing operations on the linear elements in ArcGIS software: sequentially performing element conversion and line segmentation at the element conversion points to achieve even segmentation of the original line segments by length; then repeating the above operations on the segmented line segments, determining the number of repetitions N according to the density of the required sampling viewpoints, and obtaining 2N equally divided and equally spaced sampling viewpoints; The point range directly draws and adds the determined sampling viewpoints as point elements.

5. The digital evaluation and analysis method of the urban landscape viewing system based on color coding according to claim 3 is characterized in that: Step S23 includes: S231, extracting natural landscape elements, wherein the natural landscape elements include mountain landscape; including: S2311, converting terrain raster data into vector line data with elevation information, and then generating closed contour surface feature data through the feature-to-surface tool; S2312, delineate the scope of natural landscape elements based on the landscape AOI scope information; S2313, cutting contour lines within the range of the natural landscape elements as natural landscape element data; S2314, segmentation of natural landscape elements and terrain, and fusion and update of terrain raster data; S232, extracting artificial landscape elements, wherein the artificial landscape elements include landmark buildings; including: S2321. Filter building outlines according to their geometric properties; S2322. Use reverse geocoding to obtain building functions and filter out the outlines of buildings used as commercial offices and public buildings; S2323. Use POI, AOI and collected information, supplemented by tourism software APP to select the outlines of historical and cultural landmark buildings; S2324. After verifying the popularity of the selected building outlines, they will be used as landmark architectural elements.

6. The digital evaluation and analysis method of the urban landscape viewing system based on color coding according to claim 3 is characterized in that: The S3 includes: S31. Constructing geographic environment, including terrain raster data, road element data and block data; Load the terrain raster data of the base map; Loading road feature data: Using the buffer tool, buffer the road line features according to the width value determined by the road level to generate road surface feature data; Loading block data: Generate block faces from the enclosed area of ​​the closed road line features through the feature-to-face tool, and establish a block ID attribute field to uniquely identify each block face; through the spatial connection of ArcGIS software, assign attribute information of the block ID where the building outline data is located according to the spatial position; S32. Generate background building environment: Erasing the outline of the landmark building from the original building outline data as the outline data of the background building, importing the outline data of the background building into the scene, automatically stretching the building block according to the attribute value of the height attribute in the building outline vector data, and generating the background building environment; S33, importing viewing objects: uniformly encoding the natural landscape elements and artificial landscape element data with attribute information; S34. Create other required visual elements: S35, loading terrain grid data according to terrain layers, loading the road surface element data and block surface element data, and loading the background building environment, other required visual elements and viewing objects in sequence, and then vertically projecting them onto the terrain grid graphics of the geographical environment so as to fit closely to the terrain, thereby constructing the viewing object background environment.

7. The digital evaluation and analysis method of the urban landscape viewing system based on color coding according to claim 1 is characterized in that: The step S4 comprises: S41, compiling a color coding function through CGA rules and generating a three-dimensional information model of landscape elements; S411. Setting attribute parameters of the current element in the CGA rule file; S412, converting the ID value consistent with the attribute name of the landscape element data into an HVS color value; S413, coloring the model according to the HVS color value; S414, constructing an information model generation function: the landscape element generates a three-dimensional model of the landscape element by establishing two-layer models of the main layer and the identification layer. The main layer is the main part of the model of the landscape element. After being stretched to the building height by the stretching function extrude, the three-dimensional model body is assigned a corresponding color by color coding information corresponding to the ID attribute; the identification layer is stretched 2 units high on the basis of the main layer, and the two layers of the identification layer are respectively assigned black as the identification color and white as the background color; S42, setting background color for background model: setting the color of other building outline three-dimensional models except the three-dimensional model of landscape elements to white, and the area covered by shadow to gray.

8. The digital evaluation and analysis method of the urban landscape viewing system based on color coding according to claim 1 is characterized in that: The step S5 comprises: S51, picking up sampling points in numerical order; S52, setting the camera lens to set the elevation angle, horizontal field of view and height parameters according to the general observation characteristics of the average height of a person; S53, at the sampling point, according to the lens rotation along the y-axis 0°, 90°, 180°, 270°, respectively intercept the scene photos, and name them in a unified letter identification, viewpoint ID, and viewing angle content form in a unified "_" connection method, and save them to the specified folder.

9. The digital evaluation and analysis method of the urban landscape viewing system based on color coding according to claim 1 is characterized in that: The step S6 comprises: S61, using the color of the logo layer as ROI to identify the logo layer in the image; S62, taking the background building and the environment color as ROI, obtaining a distinguished binary background image; S63, calculating the element main layer image area; S64. Calculate the color of the main layer of the feature and parse the ID value.

10. The digital evaluation and analysis method of the urban landscape viewing system based on color coding according to claim 1 is characterized in that: The step S7 includes regional landscape visibility assessment, regional landscape visibility completeness assessment and landscape element orientation analysis.

Images