CN113343781B - City functional area identification method using remote sensing data and taxi track data - Google Patents
City functional area identification method using remote sensing data and taxi track data Download PDFInfo
- Publication number
- CN113343781B CN113343781B CN202110536051.6A CN202110536051A CN113343781B CN 113343781 B CN113343781 B CN 113343781B CN 202110536051 A CN202110536051 A CN 202110536051A CN 113343781 B CN113343781 B CN 113343781B
- Authority
- CN
- China
- Prior art keywords
- analysis unit
- image
- time sequence
- data
- remote sensing
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/24323—Tree-organised classifiers
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Navigation (AREA)
- Traffic Control Systems (AREA)
- Image Processing (AREA)
Abstract
The invention relates to a city functional area identification method comprehensively using remote sensing data and taxi track data. And identifying the urban functional area by extracting the weighted mean visual characteristic of the high-spatial-resolution remote sensing image and the time sequence statistical characteristic of the data of the taxi track on-off points. When the visual features of the remote sensing image are extracted, the unbalance problem of the sample can be effectively relieved and the precision is improved by calculating the mean value weight of each basic analysis unit; the visual characteristic of the high-score remote sensing image and the statistical characteristic of taxi track data are comprehensively used, the advantages of the visual characteristic and the statistical characteristic are fully exerted, the attribute information of urban land is represented from two aspects of natural attributes and social attributes, the urban functional areas can be effectively classified, and compared with a method only using the visual characteristic or the time sequence statistical characteristic, the method provided by the invention has the advantages that the three indexes of accuracy, recall rate and F1-score are higher, and the identification accuracy of the urban functional areas is higher.
Description
Technical Field
The invention belongs to the technical field of urban geography, and particularly relates to an urban functional area identification method comprehensively using remote sensing data and taxi track data.
Background
Urban land is a general term of land with certain purposes and functions within an urban planning area, and comprises four land attributes including a natural attribute, a social attribute, an economic attribute, a legal attribute and the like. The urban spatial structure mainly comprises urban natural elements, spatial distribution patterns of social elements and mutual relations among the elements. The research on the urban spatial structure morphology appears earlier in western society, and the theoretical research as a system is gradually generated and developed under the push of urban planning practice. In 1923, by taking chicago as an example, a concentric circle mode of a spatial organization mode of city development and land use was created based on concepts of invasion and inheritance of social ecology by bogies. In 1932, babu introduced the radiation effect of traffic axis, and modified the concentric circle mode into star-shaped annular mode, so that it is closer to the real situation of single-center small and medium-scale cities. However, in China before the 80 s in the 20 th century, the urban spatial structure and form only have sporadic research results, and the research in the true sense is started after the 80 s. Since the reform is open, the economic development of China is rapid, the scale of cities is continuously expanded, the population of cities is soared, a series of urban diseases appear, the root of the urban diseases lies in the detuning of various relations among people, nature, people, spirit and substances in the process of urbanization, and how to reasonably plan and manage the spatial structure of the cities becomes a problem to be solved urgently.
At present, the land utilization status map compiled by each level of land management departments is mainly made according to remote sensing image interpretation, internal industry analysis of construction land approval information, field investigation and verification results and the like, so that the consumption of manpower and material resources is high, and the updating period is long. The remote sensing image is used for carrying out the research of land use type classification, the research focuses on extracting spectral features and textural features based on pixels at first, then scholars add structural information of target ground objects, the remote sensing image classification is carried out based on an object-oriented idea, context semantic information of images is introduced recently, and the image classification is carried out from the perspective of scene understanding. However, the above studies only reflect the natural geographic environment of a city, that is, the spatial composition and spatial arrangement of city constituent elements, most human activities are performed in a building, and the socioeconomic functional attributes of indoor activity areas cannot be determined only by remote sensing images. In recent years, the emergence of urban resident space-time behavior data such as taxi track data, social media data and the like provides urban resident continuous economic and social activity information with large sample amount, and the urban resident continuous economic and social activity information can be used for researching individual activity patterns and providing social economic environment information for urban land utilization type space distribution patterns in macroscopic forms. The taxi track data has the advantages of large data volume, high coverage rate, low cost, high updating speed and the like, and is widely applied to the research fields of traffic state evaluation, moving mode detection, hot spot area extraction, road network updating, land utilization classification and the like.
Disclosure of Invention
Aiming at the defects of the prior art, the invention provides a city functional area identification method by comprehensively using remote sensing data and taxi track data, and the city functional area is identified by extracting the weighted mean value visual characteristic of a high-spatial-resolution remote sensing image and the time sequence statistical characteristic of taxi track upper and lower point data.
In order to achieve the purpose, the technical scheme provided by the invention is an urban functional area identification method comprehensively using remote sensing data and taxi track data, and the method comprises the following steps of:
step 1, establishing a basic analysis unit of urban land based on OpenStreetMap (OSM) road data, noting the type attribute of a construction land of the basic analysis unit, and generating a land utilization map, wherein the method comprises the following substeps:
step 1.1, downloading a shape file of an OSM city base map and related attribute data, including information of urban roads, buildings, natural land, important places, railways, water systems and the like, and cutting out a research area according to an administrative boundary vector diagram of the area;
step 1.2, extracting road center lines of regional urban road data in OSM, referring to a track path generated by gathering taxi track point data, performing topology inspection and correction, and generating a regional road network vector map layer;
step 1.3, generating a basic analysis unit surface layer file by utilizing the regional road network vector layer cutting region administrative boundary vector diagram obtained in the step 1.2;
step 1.4, referring to actual city construction land types of a Baidu online map and a Gaode online map, adding construction land type attributes to corresponding basic analysis units to generate a land utilization map, wherein regions with high land utilization mixing processes are not considered in the research;
step 2.1, preprocessing the original remote sensing image data of the satellite, including geometric correction, image fusion, image mosaic and image cutting;
2.2, 50184 image objects are obtained by sequentially using three multilevel image segmentation methods of chessboard segmentation, multi-scale segmentation and spectral difference segmentation based on the high-spatial-resolution remote sensing image and the road vector diagram;
step 2.3, extracting 19 visual features of each image object including spectral features, textural features and geometric features to obtain a 19-dimensional visual feature vector of each image object;
step 2.4, taking each basic analysis unit as a unit, and obtaining a 19-dimensional weighted average visual feature vector of each basic analysis unit by calculating a weighted average value of the visual feature values of all image objects contained in each basic analysis unit;
step 3.1, preprocessing the taxi track data of one month, including coordinate system deviation correction, research area cutting and abnormal point cleaning;
step 3.2, extracting getting-on and getting-off points from the preprocessed taxi track data in the step 3.1;
step 3.3, matching the image layers of the upper and lower bus points extracted in the step 3.2 with the image layers of the basic unit surface in the step 1.3;
step 3.4, extracting 48-dimensional time sequence statistical characteristics of the data of the taxi getting-on and getting-off points in one month in each basic analysis unit;
and 4, comprehensively using the visual characteristics of the high-spatial-resolution remote sensing image and the time sequence statistical characteristics of the data of the taxi on/off points of the taxi track to identify the urban functional area.
In addition, in the step 1.2, the topology check can be performed by using a 'topology' tool in ArcMap software, the topology correction can be performed by manually correcting the road center line of the regional urban road data extracted from the OSM according to the track path of the taxi, the taxi track data is acquired by receiving information transmitted by a navigation satellite by a GPS (global positioning system) device arranged on the taxi, the sampling time interval is more than 40s, and each record contains attribute information such as a license plate number, time, longitude and latitude, direction, speed, ACC state, passenger carrying state and the like.
In step 2.4, the 19-dimensional weighted average visual feature vector of each basic analysis unit is calculated as follows:
in the formula (I), the compound is shown in the specification,
19-dimensional weighted mean visual feature vector, L, for the ith elementary analysis unitijIs 19-dimensional visual feature vector of jth image object in ith basic analysis unit, m is the number of image objects in ith basic analysis unit, piFor the weight value of each image object in the ith basic analysis unit, the calculation method is as follows:
in the formula, niRepresenting the number of image objects in the ith elementary analysis unit and N representing the number of total image objects in the investigation region.
And in step 3.3, the basic analysis unit surface image layer in step 1.3 needs to be converted into a line image layer, 4 geometric attributes of a line segment starting point, an end point, a middle point and a direction are added to the line image layer, then the line image layer is matched with the point image layer according to the nearest distance, and finally, the track points are matched into the basic analysis units on two sides of the line according to the direction attribute of the point image layer and the direction attribute of the basic analysis unit surface image layer, so that the point image layer is matched with the basic analysis unit surface image layer.
Furthermore, in step 3.4, the total amount of the getting-on and getting-off points of the taxi in 24 hours per hour per day in one month of each basic analysis unit needs to be counted first, so as to form 24-dimensional time sequence statistical feature vectors of the getting-on points and 24-dimensional time sequence statistical feature vectors of the getting-off points per day, then the 24-dimensional time sequence statistical features are divided into two categories, namely, working days including monday to friday 5 days, weekends including saturday and sunday, 48-dimensional time sequence statistical feature vectors of the getting-on points are formed by 24-dimensional average time sequence statistical feature vectors of working days and 24-dimensional average time sequence statistical feature vectors of holidays, 48-dimensional time sequence statistical feature vectors of the getting-off points are formed by 24-dimensional average time sequence statistical feature vectors of working days and 24-dimensional average time sequence statistical feature vectors of holidays, and the calculation method of the getting-on and getting-off points of each basic analysis unit is as follows:
in the formula, VonFor 24-dimensional average time sequence statistical characteristic vector of the vehicle-entering point of each basic analysis unit in working day, ShIs a taxi track experimental data set within one month, dkFor the (k) th work day,
a 24-dimensional time sequence statistical feature vector, W, of the upper vehicle point of the kth working daydIs the total number of working days in a month;
the method for calculating the 24-dimensional average time sequence statistical characteristic vector of the upper vehicle point of each basic analysis unit on the weekend is as above, correspondingly converting the 24-dimensional time sequence statistical characteristic vector of the upper vehicle point on the kth weekend into the 24-dimensional time sequence statistical characteristic vector of the upper vehicle point on the kth weekend, and converting the total days of the working day into the total days of the weekend;
in the formula, VoffFor 24-dimensional average time sequence statistical characteristic vector of departure point of each basic analysis unit in working day, ShIs a taxi track experimental data set within one month, dkFor the (k) th work day,
counting the characteristic vector W of the number of the alighting points of the kth working day in 24-dimensional time sequencedIs one monthTotal number of working days in;
the method for calculating the getting-off point 24-dimensional average time sequence statistical characteristic vector of each basic analysis unit on weekends is as above, and correspondingly, the getting-off point 24-dimensional time sequence statistical characteristic vector of the kth working day is converted into the getting-off point 24-dimensional time sequence statistical characteristic vector of the kth weekend, and the total days of the working day is converted into the total days of the weekend.
And step 4, inputting the various features extracted in step 2 and step 3 into a random forest classifier to classify the city land utilization types, identifying the functional areas of the city, wherein the land utilization types comprise five types, namely entertainment and fitness land, industrial land, public management and public service facility land, residential land and commercial land, and performing precision evaluation by taking the land utilization map generated in step 1 as a reference map and a sample set.
Compared with the prior art, the invention has the following advantages: (1) when the visual features of the remote sensing image are extracted, the mean weight of each basic analysis unit is calculated, so that the problem of unbalance of the sample can be effectively relieved, and the precision is improved; and (2) comprehensively using the visual characteristics of the high-resolution remote sensing image and the statistical characteristics of taxi track data, fully playing the advantages of the visual characteristics and the statistical characteristics, representing the attribute information of urban land from two aspects of natural attributes and social attributes, and effectively classifying urban functional areas.
Drawings
FIG. 1 is a flow chart of an embodiment of the present invention.
FIG. 2 is a schematic diagram of the spatial distribution of the basic analysis unit according to the embodiment of the present invention.
Detailed Description
The invention provides a city functional area identification method comprehensively using remote sensing data and taxi track data, which is used for identifying city functional areas by extracting weighted mean visual features of high-spatial-resolution remote sensing images and time sequence statistical features of taxi track boarding and disembarking point data.
The technical solution of the present invention is further explained with reference to the drawings and the embodiments.
As shown in fig. 1, the process of the embodiment of the present invention includes the following steps:
step 1, establishing a basic analysis unit of urban land based on OpenStreetMap (OSM) road data, noting the type attribute of a construction land of the basic analysis unit, and generating a land utilization map, wherein the method comprises the following substeps:
step 1.1, downloading the SHApefile of the OSM city base map and related attribute data, including information of urban roads, buildings, natural land, important places, railways, water systems and the like, and cutting out a research area according to the regional administrative boundary vector diagram.
Step 1.2, extracting the road center line of regional urban road data in the OSM by using ArcGIS software, referring to a track path generated by gathering taxi track point data, carrying out topology inspection and correction, and generating a regional road network vector map layer. The taxi track data is acquired by receiving information transmitted by a navigation satellite by a GPS device arranged on a taxi, the sampling time interval is more than 40s, and each record comprises attribute information such as license plate number, time, longitude and latitude, direction, speed, ACC state, passenger carrying state and the like.
And step 1.3, generating a basic analysis unit surface layer file by utilizing the regional road network vector layer cutting region administrative boundary vector diagram obtained in the step 1.2 in ArcGIS software.
And step 1.4, referring to actual types of urban construction land of the Baidu online map and the Gaode online map, adding a construction land type attribute to a corresponding basic analysis unit, and generating a land utilization map, wherein regions with high land utilization mixing process are not considered in the research.
and 2.1, preprocessing the original remote sensing image data of the satellite, including geometric correction, image fusion, image mosaic and image cutting.
And 2.2, 50184 image objects are obtained by sequentially using three multilevel image segmentation methods of chessboard segmentation, multi-scale segmentation and spectral difference segmentation based on the high-spatial-resolution remote sensing image and the road vector diagram.
And 2.3, extracting 19 visual features of each image object including spectral features, textural features and geometric features to obtain a 19-dimensional visual feature vector of each image object.
Step 2.4, taking each basic analysis unit as a unit, calculating a weighted average value of the visual characteristic values of all the image objects contained in the basic analysis unit to obtain a 19-dimensional weighted average visual characteristic vector of each basic analysis unit, wherein the calculation method comprises the following steps:
in the formula (I), the compound is shown in the specification,
19-dimensional weighted mean visual feature vector, L, for the ith elementary analysis unitjIs 19-dimensional visual feature vector of jth image object in ith basic analysis unit, m is the number of image objects in ith basic analysis unit, piFor the weight value of each image object in the ith basic analysis unit, the calculation method is as follows:
in the formula, niRepresenting the number of image objects in the ith elementary analysis unit and N representing the number of total image objects in the investigation region.
and 3.1, preprocessing the taxi track data of one month, including coordinate system deviation correction, research area cutting and abnormal point cleaning.
And 3.2, extracting getting-on and getting-off points from the preprocessed taxi track data in the step 3.1.
And 3.3, matching the upper and lower vehicle point image layers extracted in the step 3.2 with the basic unit surface image layer in the step 1.3. Firstly converting the basic analysis unit surface image layer in the step 1.3 into a line image layer, adding 4 geometric attributes of a line segment starting point, an end point, a middle point and a direction for the line image layer, then matching the line image layer with the point image layer according to the nearest distance, and finally matching the track points into the basic analysis units at two sides of the line according to the direction attribute of the point image layer and the direction attribute of the basic analysis unit surface line layer to realize the matching of the point image layer and the basic analysis unit surface image layer.
And 3.4, extracting 48-dimensional time sequence statistical characteristics of the data of the taxi getting-on and getting-off points in one month in each basic analysis unit. The method comprises the steps of firstly counting the total amount of 24-dimensional time sequence statistical feature vectors of getting-on points and getting-off points of a taxi hired every 24 hours every day in each basic analysis unit to form 24-dimensional time sequence statistical feature vectors of the getting-on points and the getting-off points every day, then dividing the 24-dimensional time sequence statistical features into two categories of working days and weekends, wherein the working days comprise Monday to Friday, the weekends comprise Saturday and Sunday, the 48-dimensional time sequence statistical feature vectors of the getting-on points are formed by 24-dimensional average time sequence statistical feature vectors of the working days and 24-dimensional average time sequence statistical feature vectors of the rest days, and the 48-dimensional time sequence statistical feature vectors of the getting-off points are formed by 24-dimensional average time sequence statistical feature vectors of the working days and 24-dimensional average time sequence statistical feature vectors of the rest days. The calculation method of the upper and lower vehicle point statistics of each basic analysis unit is as follows:
in the formula, VonFor 24-dimensional average time sequence statistical characteristic vector of the vehicle-entering point of each basic analysis unit in working day, ShIs a taxi track experimental data set within one month, dkFor the (k) th work day,
a 24-dimensional time sequence statistical feature vector, W, of the upper vehicle point of the kth working daydIs the total number of working days in a month.
The method for calculating the 24-dimensional average time sequence statistical feature vector of the upper vehicle point of each basic analysis unit on weekends is as above, and the 24-dimensional time sequence statistical feature vector of the upper vehicle point on the kth weekend is correspondingly converted into the 24-dimensional time sequence statistical feature vector of the upper vehicle point on the kth weekend, and the total days of the weekend is converted into the total days of the weekend.
In the formula, VoffFor 24-dimensional average time sequence statistical characteristic vector of departure point of each basic analysis unit in working day, ShIs a taxi track experimental data set within one month, dkFor the (k) th work day,
counting the characteristic vector W of the number of the alighting points of the kth working day in 24-dimensional time sequencedIs the total number of working days in a month.
The method for calculating the getting-off point 24-dimensional average time sequence statistical characteristic vector of each basic analysis unit on weekends is as above, and correspondingly, the getting-off point 24-dimensional time sequence statistical characteristic vector of the kth working day is converted into the getting-off point 24-dimensional time sequence statistical characteristic vector of the kth weekend, and the total days of the working day is converted into the total days of the weekend.
And 4, comprehensively using the visual characteristics of the high-spatial-resolution remote sensing image and the time sequence statistical characteristics of the data of the taxi track getting-on and getting-off points to identify the urban functional area. Inputting the various characteristics extracted in the step 2 and the step 3 into a random forest classifier to classify urban land utilization types, identifying functional areas of the city, wherein the land utilization types comprise five types, namely entertainment sports land, industrial land, public management and public service facility land, residential land and commercial land, and performing precision evaluation by taking the land utilization map generated in the step 1 as a reference map and a sample set.
The technical scheme of the invention is further illustrated by taking the experimental data of the Wuhan city Wuchang district as an example, and the effectiveness of the method is illustrated by quantitative precision evaluation.
Step 1, establishing a basic analysis unit of urban land based on OpenStreetMap road data, as shown in FIG. 2.
And 2, preprocessing the Pleiades satellite image, performing experiments for multiple times by using a multi-level image segmentation module of the image software easy to be recovered to obtain optimal segmentation parameters, and segmenting the satellite image and the road network vector layer to obtain 50184 image objects. The visual characteristics of the image object in each basic analysis unit are extracted, and the mean weight values of 19 visual characteristics are calculated by taking each basic analysis unit as a unit, and the various visual characteristics used in this example are shown in table 1.
TABLE 1 visual characteristics of high resolution remote sensing images
And 3, calculating the time sequence statistical characteristics of the getting-on and getting-off points of the taxi track data in each basic analysis unit, wherein each unit corresponds to a 48-dimensional vector.
And 4, fusing the mean weighted visual features of the remote sensing images and the timing sequence statistical features of the taxi tracks by using a random forest classifier, and classifying the urban functional areas.
To verify the effectiveness of the proposed method, comparative tests and quantitative accuracy evaluations were carried out. And respectively inputting the mean weighted visual features of the remote sensing images and the time sequence statistical features of the taxi tracks into a random forest classifier, and classifying the urban functional areas to serve as comparison experiments. For the accuracy evaluation, the following (5) to (7) were calculated using the accuracy (Precision), Recall (Recall) and F1-Score (F1-Score) as evaluation criteria.
The results of the precision evaluation of the comparative experiment are shown in table 2.
In the formula, Precision represents an accuracy index of an experimental result, TP represents the number of samples determined as positive and actually positive, and FP represents the number of samples determined as positive and actually negative.
In the formula, Recall represents a Recall index of the experimental result, TP represents the number of samples determined as positive and actually positive, and FN represents the number of samples determined as negative and actually negative.
In the formula, F1-Score represents an index of F1 Score of the experimental results.
Table 2 precision evaluation results of comparative experiments
As can be seen from table 2, the accuracy, the recall rate and the F1-score obtained by the classification method comprehensively using the mean weighted visual feature fused with the remote sensing image and the time sequence statistical feature of the taxi track are all higher than those obtained by the method using the visual feature or the time sequence statistical feature alone, which proves that the classification method provided by the invention has higher identification accuracy for the urban functional area.
In specific implementation, the above process can adopt computer software technology to realize automatic operation process.
The specific embodiments described herein are merely illustrative of the spirit of the invention. Various modifications or additions may be made to the described embodiments or alternatives may be employed by those skilled in the art without departing from the spirit or ambit of the invention as defined in the appended claims.
Claims (10)
1. A city functional area identification method comprehensively using remote sensing data and taxi track data is characterized by comprising the following steps:
step 1, establishing a basic analysis unit of urban land based on OpenStreetMap road data, noting the type attribute of a construction land of the basic analysis unit, and generating a land utilization map, wherein the method comprises the following substeps:
step 1.1, downloading a shape file of an OSM city base map and related attribute data, including urban roads, buildings, natural land, important places, railways and water system information, and cutting out a research area according to a regional administrative boundary vector diagram;
step 1.2, extracting road center lines of regional urban road data in OSM, referring to a track path generated by gathering taxi track point data, performing topology inspection and correction, and generating a regional road network vector map layer;
step 1.3, generating a basic analysis unit surface layer file by utilizing the regional road network vector layer cutting region administrative boundary vector diagram obtained in the step 1.2;
step 1.4, referring to the actual type of the urban construction land, adding the type attribute of the construction land for the corresponding basic analysis unit, and generating a land utilization map;
step 2, extracting visual characteristics of the weighted mean value based on the remote sensing image with high spatial resolution, comprising the following substeps:
step 2.1, preprocessing the original remote sensing image data of the satellite;
2.2, obtaining M image objects based on the high-spatial-resolution remote sensing image and the road vector diagram by using a multi-level image segmentation method;
step 2.3, lambda visual characteristics including spectral characteristics, textural characteristics and geometric characteristics of each image object are extracted to obtain a lambda-dimensional visual characteristic vector of each image object;
step 2.4, taking each basic analysis unit as a unit, and solving a weighted average value of the visual characteristic values of all image objects contained in each basic analysis unit to obtain a lambda-dimensional weighted average visual characteristic vector of each basic analysis unit;
step 3, extracting the time sequence statistical characteristics of the data of the getting-on and getting-off points based on the taxi track data, and comprising the following substeps:
step 3.1, preprocessing taxi track data in a certain time period;
step 3.2, extracting getting-on and getting-off points from the preprocessed taxi track data in the step 3.1;
step 3.3, matching the image layers of the upper and lower bus points extracted in the step 3.2 with the image layers of the basic unit surface in the step 1.3;
step 3.4, extracting 48-dimensional time sequence statistical characteristics of taxi getting-on and getting-off point data in each basic analysis unit within a certain time period;
and 4, comprehensively using the visual characteristics of the high-spatial-resolution remote sensing image and the time sequence statistical characteristics of the data of the taxi track getting-on and getting-off points to identify the urban functional area.
2. The method for recognizing an urban functional area by comprehensively using remote sensing data and taxi track data according to claim 1, wherein the method comprises the steps of: the topology check in the step 1.2 can be performed by using a 'topology' tool in ArcMap software, the topology correction can be performed by manually correcting the road center line of the regional urban road data extracted from the OSM according to the track path of the taxi, the taxi track data is acquired by receiving information transmitted by a navigation satellite by a GPS (global positioning system) device arranged on the taxi, the sampling time interval is more than 40s, and each record comprises a license plate number, time, longitude and latitude, direction, speed, ACC state and passenger carrying state attribute information.
3. The method for recognizing an urban functional area by comprehensively using remote sensing data and taxi track data according to claim 1, wherein the method comprises the steps of: the preprocessing in the step 2.1 comprises geometric correction, image fusion, image mosaic and image cutting.
4. The method for recognizing an urban functional area by comprehensively using remote sensing data and taxi track data according to claim 1, wherein the method comprises the steps of: and 2.2, obtaining M image objects based on the high-spatial-resolution remote sensing image and the road vector diagram by sequentially using three multilevel image segmentation methods of chessboard segmentation, multi-scale segmentation and spectral difference segmentation.
5. The method for recognizing an urban functional area by comprehensively using remote sensing data and taxi track data according to claim 1, wherein the method comprises the steps of: the visual characteristics in the step 2.3 comprise normalized difference vegetation indexes, normalized difference water body indexes, soil adjustment vegetation indexes, gray level co-occurrence matrix mean values, entropies, contrasts, correlations, non-similarities, mean values, standard deviations, skewness, areas, length-width ratios, boundary indexes, compactness, ellipse fitting, rectangle fitting, roundness and shape indexes.
6. The method for recognizing an urban functional area by comprehensively using remote sensing data and taxi track data according to claim 1, wherein the method comprises the steps of: in the step 2.4, the λ -dimensional weighted average visual feature vector of each basic analysis unit is calculated as follows:
in the formula (I), the compound is shown in the specification,
is the lambda-dimensional weighted mean visual feature vector, L, of the ith elementary analysis unitijIs the lambda-dimensional visual feature vector of the jth image object in the ith basic analysis unit, m is the number of image objects in the ith basic analysis unit, piFor the weight value of each image object in the ith basic analysis unit, the calculation method is as follows:
in the formula, niRepresenting the number of image objects in the ith elementary analysis unit and N representing the number of total image objects in the investigation region.
7. The method for recognizing an urban functional area by comprehensively using remote sensing data and taxi track data according to claim 1, wherein the method comprises the steps of: the preprocessing in the step 3.1 comprises coordinate system correction, research area cutting and abnormal point cleaning.
8. The method for recognizing an urban functional area by comprehensively using remote sensing data and taxi track data according to claim 1, wherein the method comprises the steps of: in the step 3.3, the basic analysis unit surface image layer in the step 1.3 needs to be converted into a line image layer, 4 geometric attributes including a line segment starting point, a line segment ending point, a line segment middle point and a line segment direction are added to the line image layer, then the line image layer is matched with the line image layer according to the nearest neighbor distance, and finally the track points are matched into the basic analysis units on two sides of the line according to the direction attribute of the point image layer and the direction attribute of the basic analysis unit line image layer, so that the point image layer is matched with the basic analysis unit surface image layer.
9. The method for recognizing an urban functional area by comprehensively using remote sensing data and taxi track data according to claim 1, wherein the method comprises the steps of: step 3.4, the total amount of the getting-on and getting-off points of the taxi in each 24 hours and each hour every day in a certain time period of each basic analysis unit needs to be counted first, 24-dimensional time sequence statistical feature vectors of the getting-on points and 24-dimensional time sequence statistical feature vectors of the getting-off points every day are formed, then the 24-dimensional time sequence statistical features are divided into two types of working days and weekends, the working days comprise Monday to Friday, 5 days, the weekends comprise Saturday and Sunday, the 48-dimensional time sequence statistical feature vectors of the getting-on points are formed by 24-dimensional average time sequence statistical feature vectors of the working days and 24-dimensional average time sequence statistical feature vectors of the rest days, and the calculation method of the statistics of the getting-on and getting-off points of each basic analysis unit comprises the following steps:
in the formula, VonA 24-dimensional average time sequence statistical characteristic vector S of the upper vehicle point of each basic analysis unit in working dayshIs a taxi track experimental data set within a certain period of time, dkFor the (k) th work day,
a 24-dimensional time sequence statistical feature vector, W, of the upper vehicle point of the kth working daydIs the total number of working days in a certain period of time;
the method for calculating the 24-dimensional average time sequence statistical characteristic vector of the upper vehicle point of each basic analysis unit on the weekend is as above, correspondingly converting the 24-dimensional time sequence statistical characteristic vector of the upper vehicle point on the kth weekend into the 24-dimensional time sequence statistical characteristic vector of the upper vehicle point on the kth weekend, and converting the total days of the working day into the total days of the weekend;
in the formula, VoffA 24-dimensional average time sequence statistical characteristic vector of a get-off point of each basic analysis unit in a working day, ShIs a taxi track experimental data set within a certain period of time, dkFor the (k) th work day,
counting the characteristic vector W of the number of the alighting points of the kth working day in 24-dimensional time sequencedIs the total number of working days in a certain period of time;
the method for calculating the getting-off point 24-dimensional average time sequence statistical characteristic vector of each basic analysis unit on weekends is as above, and correspondingly, the getting-off point 24-dimensional time sequence statistical characteristic vector of the kth working day is converted into the getting-off point 24-dimensional time sequence statistical characteristic vector of the kth weekend, and the total days of the working day is converted into the total days of the weekend.
10. The method for recognizing an urban functional area by comprehensively using remote sensing data and taxi track data according to claim 1, wherein the method comprises the steps of: and 4, inputting various characteristics extracted in the steps 2 and 3 into a random forest classifier to classify the city land utilization types, identifying the functional areas of the city, wherein the land utilization types comprise five types, namely entertainment sports land, industrial land, public management and public service facility land, residential land and commercial land, and performing precision evaluation by taking the land utilization map generated in the step 1 as a reference map and a sample set.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110536051.6A CN113343781B (en) | 2021-05-17 | 2021-05-17 | City functional area identification method using remote sensing data and taxi track data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110536051.6A CN113343781B (en) | 2021-05-17 | 2021-05-17 | City functional area identification method using remote sensing data and taxi track data |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113343781A CN113343781A (en) | 2021-09-03 |
| CN113343781B true CN113343781B (en) | 2022-02-01 |
Family
ID=77469017
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110536051.6A Active CN113343781B (en) | 2021-05-17 | 2021-05-17 | City functional area identification method using remote sensing data and taxi track data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113343781B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106991142A (en) * | 2017-03-22 | 2017-07-28 | 湖州师范学院 | A kind of method that urban function region is recognized based on wechat data and interest point data |
| CN109978040A (en) * | 2019-03-19 | 2019-07-05 | 西南交通大学 | A kind of traffic transport power distribution forecasting method |
| CN110298500A (en) * | 2019-06-19 | 2019-10-01 | 大连理工大学 | A kind of urban transportation track data set creation method based on taxi car data and city road network |
| CN111651502A (en) * | 2020-06-01 | 2020-09-11 | 中南大学 | City functional area identification method based on multi-subspace model |
| CN111814596A (en) * | 2020-06-20 | 2020-10-23 | 南通大学 | Automatic city function partitioning method for fusing remote sensing image and taxi track |
-
2021
- 2021-05-17 CN CN202110536051.6A patent/CN113343781B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106991142A (en) * | 2017-03-22 | 2017-07-28 | 湖州师范学院 | A kind of method that urban function region is recognized based on wechat data and interest point data |
| CN109978040A (en) * | 2019-03-19 | 2019-07-05 | 西南交通大学 | A kind of traffic transport power distribution forecasting method |
| CN110298500A (en) * | 2019-06-19 | 2019-10-01 | 大连理工大学 | A kind of urban transportation track data set creation method based on taxi car data and city road network |
| CN111651502A (en) * | 2020-06-01 | 2020-09-11 | 中南大学 | City functional area identification method based on multi-subspace model |
| CN111814596A (en) * | 2020-06-20 | 2020-10-23 | 南通大学 | Automatic city function partitioning method for fusing remote sensing image and taxi track |
Non-Patent Citations (2)
| Title |
|---|
| Recognition of Urban Functions and Mixed Use Based on Residents’ Movement and Topic Generation Model: The Case of Wuhan, China;Haifu Cui等;《Remote Sensing 2020》;20200906;第1-21页 * |
| 融合GPS轨迹和POI数据关联规则的城市功能区识别;冯慧芳,杨文亮;《测绘科学技术学报》;20200830;第37卷(第4期);第414-420页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113343781A (en) | 2021-09-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| AU2005232219B2 (en) | Forecasting based on geospatial modeling | |
| CN104318324B (en) | Shuttle Bus website and route planning method based on taxi GPS records | |
| CN112447041B (en) | Method and device for identifying operation behavior of vehicle and computing equipment | |
| CN111814596B (en) | Automatic city function partitioning method for fusing remote sensing image and taxi track | |
| CN109903553B (en) | Multi-source data mining bus station identification and inspection method | |
| CN110555544B (en) | Traffic demand estimation method based on GPS navigation data | |
| CN112579718B (en) | Urban land function identification method and device and terminal equipment | |
| CN110990661B (en) | Region of interest access extraction method based on road network constraint and density clustering | |
| CN107656987A (en) | A kind of subway station function method for digging based on LDA models | |
| CN114363842B (en) | Bus passenger departure station prediction method and device based on mobile phone signaling data | |
| CN105844031B (en) | A kind of urban transportation gallery recognition methods based on mobile phone location data | |
| CN110413855A (en) | A kind of region entrance Dynamic Extraction method based on taxi drop-off point | |
| CN110853156A (en) | Passenger OD identification method integrating bus GPS track and IC card data | |
| CN114662774A (en) | City block vitality prediction method, storage medium and terminal | |
| CN112215059A (en) | Urban village identification and population estimation method and system based on deep learning and computer readable storage medium | |
| CN115412857A (en) | Resident travel information prediction method | |
| CN114511432A (en) | Digital country management service system based on block chain | |
| CN113672788B (en) | Urban building function classification method based on multi-source data and weight coefficient method | |
| Ratra et al. | A Comprehensive Review on Crime Patterns and Trends Analysis using Machine Learning | |
| CN111400877B (en) | Intelligent city simulation system and method based on GIS data | |
| CN113256978A (en) | Method and system for diagnosing urban congestion area and storage medium | |
| CN112364776A (en) | Urban functional area recognition model establishing and classifying method and device | |
| CN113343781B (en) | City functional area identification method using remote sensing data and taxi track data | |
| Zhou et al. | Big data for intrametropolitan human movement studies A case study of bus commuters based on smart card data | |
| CN110610446A (en) | County town classification method based on two-step clustering thought |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |