CN111288999A - Pedestrian road network attribute detection method, device and equipment based on mobile terminal - Google Patents
Pedestrian road network attribute detection method, device and equipment based on mobile terminal Download PDFInfo
- Publication number
- CN111288999A CN111288999A CN202010104216.8A CN202010104216A CN111288999A CN 111288999 A CN111288999 A CN 111288999A CN 202010104216 A CN202010104216 A CN 202010104216A CN 111288999 A CN111288999 A CN 111288999A
- Authority
- CN
- China
- Prior art keywords
- data
- attribute
- road network
- track
- training
- 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.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
- G01C21/26—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 specially adapted for navigation in a road network
- G01C21/28—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 specially adapted for navigation in a road network with correlation of data from several navigational instruments
- G01C21/30—Map- or contour-matching
- G01C21/32—Structuring or formatting of map data
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
- G01C21/26—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 specially adapted for navigation in a road network
- G01C21/34—Route searching; Route guidance
- G01C21/3407—Route searching; Route guidance specially adapted for specific applications
- G01C21/3415—Dynamic re-routing, e.g. recalculating the route when the user deviates from calculated route or after detecting real-time traffic data or accidents
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S19/00—Satellite radio beacon positioning systems; Determining position, velocity or attitude using signals transmitted by such systems
- G01S19/38—Determining a navigation solution using signals transmitted by a satellite radio beacon positioning system
- G01S19/39—Determining a navigation solution using signals transmitted by a satellite radio beacon positioning system the satellite radio beacon positioning system transmitting time-stamped messages, e.g. GPS [Global Positioning System], GLONASS [Global Orbiting Navigation Satellite System] or GALILEO
- G01S19/42—Determining position
- G01S19/45—Determining position by combining measurements of signals from the satellite radio beacon positioning system with a supplementary measurement
Landscapes
- Engineering & Computer Science (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Computer Networks & Wireless Communication (AREA)
- Navigation (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention discloses a method, a device and equipment for detecting pedestrian road network attributes based on a mobile terminal, wherein the method comprises the following steps: the method comprises the following steps that after sensor data are collected by mobile terminals of a plurality of users to obtain original data, road attributes are recorded by one part of data to be used for constructing training set data, and the other part of data is used as track data to be used for attribute detection; carrying out classification training on sample data with different attributes by using training set data to obtain a classification model; detecting the attribute of the track data by using the classification model obtained by training, and performing data fusion with the position information to obtain GPS data with attribute information; based on the existing pedestrian road network data, the attributes detected by the track points are given to the matched position points in the road network, and the road network data with the attribute information is obtained after voting and correction processing. The invention solves the problem that the navigation system in the prior art can not provide personalized navigation service with specific requirements for users; the road attributes of the pedestrian road network are detected by adopting data collected by the smart phone sensor, and a data basis is provided for realizing personalized pedestrian navigation.
Description
Technical Field
The invention relates to the technical field of intelligent traffic data processing, in particular to a pedestrian network attribute detection method, device and equipment based on a mobile terminal.
Background
As an important component of intelligent traffic, the pedestrian road network describes a topological graph of geometric relationships of pedestrian road segments, and is basic data of a pedestrian navigation system. The pedestrian navigation system aims at planning the path of the pedestrian road network, which means that the pedestrian road network needs to be constructed and updated in time, but most of the existing methods ignore the addition of attribute information in the basic data of the pedestrian road network. During path planning, measurement of attribute information is lacked, path planning is performed only by using shortest path analysis, and special travel requirements cannot be met. The road attribute is the primary consideration of special group trips, such as travelers riding bicycles and people with limited actions, and road obstacles greatly influence the accessibility and comfort of trips. Particularly for wheelchair users, when the users go out, the users can not pass through the pedestrian overpass or stairs without matched barrier-free service facilities. In addition, when the road meets roads with different slope values, the influence on the travel accessibility is far from the other. For the existing pedestrian navigation service, attribute information is lost in basic data, so that optimal paths meeting different travel requirements cannot be provided for travel groups, and personalized navigation cannot be realized.
Namely, the navigation system in the prior art cannot provide personalized navigation service with specific requirements for users, so that the navigation system is poor in friendliness to partial users, and the use feeling of a trip group on the navigation system is reduced to a certain extent.
Disclosure of Invention
The invention aims to solve the technical problems that a method, a device and equipment for detecting the attributes of a pedestrian road network based on a mobile terminal are provided, and the problems that the attribute information is lacked in the road network basic data, a navigation system in the prior art cannot provide personalized navigation service with specific requirements for users, and the experience of travel groups on the navigation system is poor are solved; the sensor data of the mobile terminal is adopted to detect the road attributes of the pedestrian road network, the basic data of the pedestrian road network is enriched, a data base is provided for realizing personalized pedestrian navigation service, and the special requirements of different travel groups are met.
A pedestrian network attribute detection method based on a mobile terminal comprises the following steps:
the method comprises the steps that the mobile terminals of a plurality of users collect sensor data, and required sensor data are extracted after the original data are obtained; recording road network attributes of data collected by part of user mobile terminals, and using the road network attributes to construct training set data, and directly uploading data collected by the rest of user mobile terminals to a cloud end to obtain crowdsourcing data, and using the crowdsourcing data as track data for attribute detection;
preprocessing data used for constructing a training set, sampling according to a sliding window, calculating a characteristic value of each sample, marking a corresponding attribute type, screening the characteristic values by using a characteristic selection algorithm in machine learning, and constructing the training data set;
according to the obtained training data set, carrying out classification training on sample data with different attributes by adopting a machine learning method to obtain a classification model suitable for the road network attributes;
carrying out data processing on the crowdsourced trajectory data, and carrying out attribute detection by using a trained classification model; performing data fusion on the position information and the attribute information of the track data to obtain GPS data with attribute information;
map matching is carried out based on basic data of the existing pedestrian road network, and attribute information of track points is given to position points matched to the existing road network; obtaining a unique label of the position point by using a majority voting method for the position point matched with the plurality of attribute labels; and correcting the abnormal label to obtain the road network data with the attribute information.
The pedestrian network attribute detection method based on the mobile terminal comprises the steps that the mobile terminals of a plurality of users collect sensor data, and required sensor data are extracted after original data are obtained; the method comprises the following steps of recording road network attributes of data collected by part of user mobile terminals, constructing training set data, directly uploading data collected by the rest of user mobile terminals to a cloud end to obtain crowdsourcing data, and using the crowdsourcing data as track data for attribute detection:
the method comprises the steps that data of a plurality of sensors are collected through mobile terminals of a plurality of users, and data of an accelerometer, a barometer and a GPS are extracted; and obtaining triaxial acceleration, an air pressure value and GPS data, and reserving timestamp data.
The method for detecting the attributes of the pedestrian network based on the mobile terminal comprises the following steps of preprocessing data used for constructing a training set, sampling according to a sliding window, calculating a characteristic value of each sample, marking a corresponding attribute type, screening the characteristic values by using a characteristic selection algorithm in machine learning, and constructing the training data set, wherein the method comprises the following steps:
eliminating noise data in original data used for constructing a training set, and filtering an air pressure value to obtain initial data used for constructing the training set;
carrying out data sampling on the obtained data for constructing the training set through a sliding window to obtain a plurality of samples of the data set;
after data sampling is finished, selecting a mean value, a variance, a correlation coefficient and a gas pressure difference as initial characteristics of samples, calculating a characteristic value of each sample, retaining first time stamp data of each sample, screening the initial characteristics by adopting a characteristic selection function of Weka, and extracting an optimal characteristic subset;
and after the characteristic value of each sample is obtained, adding the corresponding attribute label to obtain the finally required training set data.
The method for detecting the attribute of the pedestrian network based on the mobile terminal comprises the following steps of:
determining a sampling frequency according to the acquired data;
setting a variance threshold value and a rejection amount according to the calculated sampling frequency;
removing the initial and final data according to the removal amount;
and eliminating the data collected in the non-motion state according to the set variance threshold.
The method for detecting the attribute of the pedestrian network based on the mobile terminal comprises the following steps of carrying out classification training on sample data with different attributes by adopting a machine learning method according to an obtained training data set to obtain a classification model suitable for the attribute of the pedestrian network, wherein the step of obtaining the classification model suitable for the attribute of the pedestrian network comprises the following steps:
and taking the obtained final training set data as model input, and training by taking a K-adjacent model as a classification model to obtain the classification model suitable for the road network attribute.
The pedestrian network attribute detection method based on the mobile terminal comprises the steps of carrying out data processing on crowd-sourced trajectory data and carrying out attribute detection by using a trained classification model; the step of performing data fusion on the position information and the attribute information of the track data to obtain the GPS data with the attribute information comprises the following steps:
preprocessing and sampling crowdsourced trajectory data, calculating characteristic values of all samples according to selected characteristics in training data, and meanwhile, reserving first time stamp data of each sample;
performing attribute detection on the processed track data by using the trained K-adjacent model to obtain attribute information of each sample of the track data and time information;
and performing data fusion on the obtained attribute information with the time information and the position information to obtain the GPS data with the attribute information of each track.
The method for detecting the attribute of the pedestrian network based on the mobile terminal comprises the steps of carrying out map matching based on basic data of the existing pedestrian network, and endowing attribute information of track points to position points matched to the existing pedestrian network; obtaining a unique label of the position point by using a majority voting method for the position point matched with the plurality of attribute labels; the step of correcting the abnormal label to obtain the road network data with the attribute information comprises the following steps:
acquiring basic data of the existing road network, sampling the fused track data according to a certain time window length, and performing attribute matching on each sample as a track segment according to the position;
after matching is finished, due to the difference between the road network point density and the track point density, the position points of a part of road network are matched to a plurality of attribute tags, and a plurality of attribute tags of each position point are voted by using a majority voting method to obtain a unique tag.
After the unique label is obtained, due to the influence of classification precision, the attribute detected by part of position points is wrong, logic judgment is needed, and an abnormal label is obtained and corrected to obtain the final road network basic data with the attribute.
A pedestrian network attribute detection device based on a mobile terminal comprises:
the acquisition module is used for acquiring sensor data by mobile terminals of a plurality of users, and extracting required sensor data after acquiring original data; recording road network attributes of data collected by part of user mobile terminals, and using the road network attributes to construct training set data, and directly uploading data collected by the rest of user mobile terminals to a cloud end to obtain crowdsourcing data, and using the crowdsourcing data as track data for attribute detection;
the construction module is used for preprocessing data used for constructing a training set, calculating a characteristic value of each sample according to sliding window sampling, marking a corresponding attribute type, screening the characteristic values by using a characteristic selection algorithm in machine learning, and constructing the training data set;
the classification module is used for performing classification training on sample data with different attributes by using the obtained training data set and adopting a machine learning method to obtain a classification model suitable for road network attributes;
the fusion module is used for carrying out data processing on the crowdsourced trajectory data and carrying out attribute detection by using a trained classification model; performing data fusion on the position information and the attribute information of the track data to obtain GPS data with attribute information;
the matching module is used for carrying out map matching based on basic data of the existing pedestrian road network and endowing the attribute information of the track points to the position points matched to the existing road network; obtaining a unique label of the position point by using a majority voting method for the position point matched with the plurality of attribute labels; and correcting the abnormal label to obtain the road network data with the attribute information.
A computer device comprising a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by one or more processors comprises the steps for performing the mobile terminal based pedestrian network property detection method according to any of claims 1-7.
A non-transitory computer readable storage medium, wherein instructions of the storage medium, when executed by a processor of an electronic device, enable the electronic device to perform any one of the above steps of the method for detecting a pedestrian network attribute based on a mobile terminal.
Compared with the prior art, the embodiment of the invention has the following advantages:
according to the method provided by the embodiment of the invention, the smart phone is used for collecting data of a plurality of sensors, processing the data, labeling attribute labels, and selecting proper sample characteristics for training a classification model so as to realize the function of detecting the attributes of the track data. Road attributes in the pedestrian network are automatically detected and added into basic data of the pedestrian network, the current situation that attribute factors are not considered in path planning is improved, the optimal path planning is achieved, and special requirements of pedestrian traveling are met.
The invention adopts the data collected by the smart phone sensor to detect the road attribute of the pedestrian road network, enriches the basic data of the pedestrian road network, provides a data base for realizing personalized pedestrian navigation service, and meets the special requirements of different travel groups.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments described in the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
Fig. 1 is a schematic flow chart of a method for detecting attributes of a pedestrian network based on a mobile terminal according to an embodiment of the present invention;
FIG. 2 is a schematic diagram illustrating an example of extracting slope road data of a pedestrian network attribute detection method based on a mobile terminal according to an embodiment of the present invention
Fig. 3 is a schematic diagram illustrating an example of extracting pedestrian overpass (stair) data in a pedestrian network attribute detection method based on a mobile terminal according to an embodiment of the present invention.
Fig. 4 is a functional schematic block diagram of a pedestrian network attribute detection device based on a mobile terminal according to an embodiment of the present invention.
Fig. 5 is a schematic structural diagram of a pedestrian network attribute detection device based on a mobile terminal according to a further embodiment of the present invention.
Detailed Description
In order to make the technical solutions of the present invention better understood, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The inventor finds that the existing pedestrian network basic data neglects the addition of the attribute information of the pedestrian network. In the process of path analysis, the navigation system lacks measurement of attribute information, only shortest path analysis is used for path planning, and special travel requirements cannot be met. The road attribute is the primary consideration of special group trips, such as travelers riding bicycles and people with limited actions, and road obstacles greatly influence the accessibility and comfort of trips. Particularly for wheelchair users, when the users go out, the users can not pass through the pedestrian overpass or stairs without matched barrier-free service facilities. In addition, when the road meets roads with different slope values, the influence on the travel accessibility is far from the other. Aiming at the existing pedestrian navigation service, the optimal paths meeting different travel requirements cannot be provided for travel groups, so that personalized navigation cannot be realized.
In order to solve the above problems, in the embodiment of the present invention, the attribute information of the pedestrian network is added to the basic data of the navigation system to provide a personalized navigation service with a specific requirement for a user, and an intelligent terminal, for example, a smart phone is used to collect data of a plurality of sensors, process the data and label attribute labels, and select a suitable sample feature for training a classification model, so as to implement a function of detecting an attribute of trajectory data. Road attributes in the pedestrian network are automatically detected and added into basic data of the pedestrian network, the current situation that attribute factors are not considered in path planning is improved, the optimal path planning is achieved, and special requirements of pedestrian traveling are met.
The intelligent mobile terminal such as a smart phone and the like is internally provided with various sensors including an accelerometer, a gyroscope, a magnetometer, a barometer, a photoreceptor, a GPS and the like, so that the invention adopts multi-sensor data to detect the attribute information of the pedestrian path section. Due to the common use of the smart phone, the method disclosed by the invention realizes attribute detection by using crowdsourcing data. The GPS data with the attribute information is mapped to a pedestrian road network of OpenStreetMap (OSM for short, Chinese is a public map), and the visualization of the attribute in the road network is realized. And finally, building a platform, uploading data collected by each user to the platform, obtaining attribute information of each track by adopting the method, adding the attribute information into the basic data of the pedestrian network, and considering the attributes in the path planning process to realize path planning meeting special requirements.
Various non-limiting embodiments of the present invention are described in detail below with reference to the accompanying drawings.
Referring to fig. 1, a method for detecting a pedestrian network attribute based on a mobile terminal in an embodiment of the present invention is shown. In this embodiment, the method may include, for example, the steps of:
s210: acquiring sensor data through mobile terminals of a plurality of users, and extracting required sensor data after acquiring original data; road network attributes of data collected by a part of user mobile terminals are recorded in the collected sensor data and used for constructing training set data, and data collected by the rest of user mobile terminals are directly uploaded to a cloud to obtain crowdsourcing data and used as track data for attribute detection.
In the embodiment, the data of a plurality of sensors are collected through N smart phones of a plurality of clients, and the data of three sensors, namely an accelerometer, a barometer and a GPS, are extracted; after obtaining the triaxial acceleration, the air pressure value and the GPS data, the following preprocessing is carried out on the data while the timestamp data is kept.
S220: preprocessing data used for constructing a training set, then sampling according to a sliding window, calculating a characteristic value of each sample, marking a corresponding attribute type, screening the characteristic values by using a characteristic selection algorithm in machine learning, and constructing the training data set;
wherein, the step S220 specifically includes:
s221, noise data in original data used for constructing a training set are removed, due to obvious shaking of a mobile phone in a motion state, an air pressure value cannot be directly used for representing real air pressure change, filtering processing needs to be carried out on the air pressure value, and then the original data used for constructing the training set are obtained;
because the uncertain influence factors of the walking of the pedestrian are many, noise exists in the acceleration data, for example, due to the operation of equipment, the start and the end of the acquisition process are in a static state, or due to the fact that the pedestrian stays in place due to the uncertain factors. The data collected in these non-motion states are the noise data in the data. Noise data needs to be eliminated so as to avoid influencing model training and subsequent detection precision. In addition, the vibration of the mobile phone is obvious in the motion state, and the acquired air pressure value cannot be directly used for representing the real change of the air pressure, so that the air pressure value is filtered by using a Butterworth filter.
Firstly, a sampling frequency is determined according to the acquired data, and the sampling frequency is determined by the difference delta t between adjacent times.
frequency＝1000/(∑Δt/n)
Where n is the data volume.
The variance threshold and reject amount reject are set according to the calculated sampling frequency, which is specifically shown in the following table. The variance threshold is used for judging whether the acquired data is in a motion state, namely, a path slope or a stair; the elimination quantity refers to a part of data after the beginning and before the end of the data acquisition process.
TABLE 3.1 numerical relationship between sampling frequency, variance threshold, and reject amount
Secondly, the time window size can be set to 1.25, 2.56, 5.12, 7.68, etc., the time window length time _ window is determined according to the requirement, and the size of the variance threshold std _ thre is determined by the sampling frequency. The samples are segmented according to the size of the time window, and the variance std _ z of the z-axis acceleration in each sample is calculated and compared with a variance threshold. Variance std _ z of z-axis acceleration in each sample>And when the std _ thre variance threshold value exists, the sample is considered as a motion state, and the data is labeled. After the motion state of each sample is obtained, the start and end time information of each sample is associated with the motion state for the next data extraction. Wherein, the time window is used for segmenting the data and is calculated by the following formula
And
respectively, the position of the start and end of the ith sample.
And finally, eliminating the initial and final data according to the elimination amount, comparing the original data with the motion state, extracting the data acquired by the motion state through the comparison of the time stamps, and storing the data to obtain the data finally used for constructing the training set.
As shown in fig. 2 and fig. 3, fig. 2 is a schematic diagram illustrating an example of extracting slope road data of a method for detecting attributes of a pedestrian network based on a mobile terminal according to an embodiment of the present invention
Fig. 3 is a schematic diagram illustrating an example of extracting pedestrian overpass (stair) data in a pedestrian network attribute detection method based on a mobile terminal according to an embodiment of the present invention.
And S222, carrying out data sampling on the obtained data for constructing the training set through a sliding window to obtain a plurality of samples of the data set.
In the invention, data used for constructing the training set is obtained after the data is preprocessed. And determining a sliding window slide _ window according to the sampling frequency and the time window time _ window, sampling data by taking 50% as window overlap, and using the data for the subsequent feature calculation. Wherein the sampling position of each sample is specifically determined by the following equation.
slide_window＝frequency*time_window
S223, after data sampling is completed, selecting a mean value, a variance, a correlation coefficient and a barometric pressure difference as initial characteristics of the samples, calculating characteristic values of the samples, keeping first time stamp data of the samples, screening the initial characteristics by adopting a characteristic selection function of Weka, and extracting an optimal characteristic subset
In the invention, after data preprocessing is completed, a plurality of samples of a data set are obtained, the mean value, the variance, the correlation coefficient and the air pressure difference are selected as initial characteristics of the samples, the characteristic value of each sample is calculated, the first piece of time stamp data of each sample is reserved, then the initial characteristics are screened by adopting the characteristic selection function of Weka, and the optimal characteristic subset is extracted. Weka, as a machine learning integration platform, is a common tool for data mining. One of the functions Select attributes in Weka is to search all possible attribute combinations in the data to find the attribute subset with the best prediction effect. In order to enable the classification accuracy of the sub-models to be better, the initial features of the training set are screened by adopting the Weka feature selection function, and the optimal feature subset is extracted to obtain the final training set.
The selected initial characteristics are illustrated in the following table.
TABLE 3.2 characteristic values and their description
The specific calculation formula is as follows:
Δbaro＝baroend-barobegin
where m denotes the amount of data in each sample, ai(i ═ 1,2, and 3) represent xyz triaxial acceleration data, respectively.
And S224, after the characteristic values of the samples are obtained, adding the corresponding attribute labels to obtain the finally required training set data.
In the embodiment of the invention, after the characteristic value of each sample is calculated, the corresponding attribute label is added to obtain the required training set. Road attributes are mainly classified into road gradient and road obstacle. The method is divided into three categories of flat roads, sloping roads and stairs, and the two categories are divided into different types according to the difference of the slope values. The attribute types are specifically shown in the following table.
TABLE 3.3 Attribute types and descriptions
S230: and carrying out classification training on sample data with different attributes by adopting a machine learning method according to the obtained training data set to obtain a classification model.
In one embodiment, the step S230 specifically includes:
s231, the obtained final training set data is used as model input, and a K-adjacent model is used as a classification model for training to obtain the classification model suitable for the road network attribute.
The invention relates to training and detection:
using a K-Nearest neighbor (KNN) model, classification is performed by measuring distances between different feature values, and if most of K Nearest neighbor (i.e., Nearest neighbor in the feature space) samples in the feature space belong to a certain class, then the samples also belong to the class and have the characteristics of the class samples.
The K value is selected to determine that the KNN model has a good training result to a great extent, and meanwhile, the distance between the objects is calculated to serve as a non-similarity index between the objects, so that the problem of matching between the objects is solved. The distance is calculated using the euclidean distance, and the formula is as follows:
in the formula, x and y are objects requiring the distance between the two, and M is the number of the objects in the data set.
The KNN model is determined according to the class of the nearest sample or samples rather than a single object class decision in the aspect of determining the classification decision, so that the problem that the training object is divided into multiple classes is avoided. In addition, the KNN model is relatively low in training time complexity and high in accuracy, so that the method adopts the KNN model as a classification model for training, trains a K-proximity model to obtain the classification model, and is applied to attribute detection of the track data.
S240: carrying out data processing on the crowdsourced trajectory data, and carrying out attribute detection by using a trained classification model; and performing data fusion on the position information and the attribute information of the track data to obtain the GPS data with the attribute information.
In the embodiment of the invention, the KNN model is trained by using the obtained training set, and after the better classification precision is achieved, the classification model suitable for road network attribute detection is obtained and is used for the attribute detection of the track data.
The step S240 includes:
s241, in the same preprocessing mode as the training data, eliminating noise data in the crowdsourced trajectory data, filtering the air pressure value, then sampling the data, calculating the characteristic value of each sample according to the selected characteristics in the training data, and meanwhile keeping the first time stamp data of each sample.
And S242, after the processed data are obtained, carrying out attribute detection on the processed track data by using a trained KNN model (K-adjacent model), so that attribute information of each sample of the track data is obtained, and the attribute information is provided with time information.
S243, performing data fusion on the obtained attribute information with time information and the position information to obtain GPS data with attribute information of each track;
in this step, for example, by comparing the time stamps and the length of the sampling window, the attribute tag of each sample is assigned to the track segment data corresponding to the sample, so as to realize data fusion of the attribute information and the position information, and thus the GPS data of each track can be provided with the attribute information.
In the invention, after the attribute information with the timestamp data is obtained, the attribute information and the position information are subjected to data fusion. Because the data is sampled through the sliding window, each track is divided into a plurality of track segments for feature calculation and attribute detection. Therefore, each track segment has unique attribute information but points to a plurality of position data, so that the aim of adding the road network basic data can be fulfilled only by carrying out data fusion processing. And the attribute information carries time stamp data, the time stamp of the attribute information is compared with the time stamp of the original track data, the track section GPS data pointed by a certain attribute is determined according to the length of a sampling window, and the two types of data are combined, namely the position data carries the attribute information.
S250: map matching is carried out based on basic data of the existing pedestrian road network, and attribute information of track points is given to position points matched to the existing road network; obtaining a unique label of the position point by using a majority voting method for the position point matched with the plurality of attribute labels; and after the abnormal label is corrected, the road network data with the attribute information is obtained.
Wherein, the step S250 specifically includes:
s251, firstly, acquiring basic data of the existing road network, sampling the fused track data according to a certain time window length, and performing attribute matching on each sample as a track segment according to the position. The actual road network is a connection combination of a plurality of road sections, and each road section is used as a matched candidate road section. And sequentially calculating the distance from each GPS point in the track segment to the candidate road network through the Euclidean distance, and summing to represent the distance index from the track segment to the road segment. And after the distance indexes from each track section to each candidate road section are obtained, selecting the road section with the closest distance as a matching object. After the road network sections matched with the track sections are determined, because the position points in the road network are the characteristic points of the sections, each section only has two starting points and two ending points, the number of points between the two position points of the road network section is required to be increased in a difference mode, and the accuracy of position matching is improved. According to the Euclidean distance, obtaining points of the track segment, which are closest to the matched object, and assigning the attributes of the track points to the matched position points;
s252, after matching is completed, because the road network point density after point number increase and the track point density are still different, the position points of a part of road network are matched to a plurality of attribute tags, and a plurality of attribute tags of each position point are voted by using a majority voting method to obtain a unique tag;
and S253, after the unique label is obtained, due to the influence of classification precision, the attribute detected by part of position points is wrong, logic judgment is needed, and an abnormal label is obtained and corrected to obtain the final road network basic data with the attribute.
In this embodiment, due to the problem of classification accuracy, one or a few abnormal attributes may appear in the continuous trace points with the same attribute, and the abnormal label needs to be logically determined and corrected, so as to improve the accuracy of road network attribute detection.
In the embodiment of the invention, the attribute information is obtained after the position data collected by the user is fused, but the real road network data cannot be directly reflected due to the positioning error of the GPS, so that the fused data needs to be matched based on the existing road network data, and the actual road network has the attribute information. The matching process mainly samples the track data according to a certain time window length, and each sample is used as a track segment for position matching.
First, the road network link closest to the track segment is found as a matching object. In order to avoid the deviation of GPS data during track acquisition, the error of single-point matching is caused. The invention takes each road segment of the road network as a matching candidate based on the matching of the track segment, respectively calculates the distance from the track point to the candidate road segment, namely calculates the linear distance from the point to the line segment, and accumulates the calculated distance of each point to obtain the distance index. And selecting the candidate road network corresponding to the shortest distance index as a final matching object. The distance calculation formula is as follows:
wherein xi，yiThe longitude and the latitude of the track point are respectively represented, ABC is determined By a straight line corresponding to the candidate road section, namely the straight line Ax + By + C is 0 and passes through the candidate road section.
Secondly, the matched road sections are subjected to point adding operation. The existing road network consists of all road sections, and the road sections only have GPS data of two initial position points and two final position points, and the position points of the road sections need to be increased so as to improve the matching accuracy of the positions. And determining the distance between the increasing points according to the empirical value, and then acquiring the GPS data of the increasing points in the road section by an interpolation method.
And then, performing final point attribute matching based on the road section after point addition, and assigning the attribute label of each track point of the track section to the position point of the closest point on the matched road section. And calculating the distance between each point in the track point value matching road section through the Euclidean distance, and giving the attribute to the position point with the closest distance to complete the attribute matching of the track point.
Euclidean distance:
in the formula, rho is the distance between point pairs; x, y are the longitude and latitude of the location point, respectively.
And finally, performing attribute correction operation. Due to the data acquisition frequency, it is not possible to match only a unique attribute tag to a location point of a road network. Therefore, after matching is completed, some position points in the road network may be matched to multiple attribute labels, and these position points need to be processed to obtain a unique label. The invention adopts a majority voting method to select the main elements. The majority voting method is simple and easy to implement, and the implementation speed is high. The majority voting method mainly finds out most elements in a given unordered array, and the occurrence frequency of the most elements needs to exceed 50%. And scanning and judging a plurality of attribute tags of each position point to obtain a plurality of elements of the attribute tags as final attribute tags of the position point. Due to detection error problems, one or a few different attribute tags may occasionally appear in a track of the same attribute. However, the point attribute may be used as the attribute of the link only when a certain number of point attributes are required to be continuously present. Therefore, the attribute label is used as an abnormal label, logic judgment is carried out, the abnormal label is corrected to be a normal label, and the final detection precision is improved.
Therefore, the road attribute detection method based on the smart phone provided by the invention provides a method for automatically detecting the road attribute of a road network by using data acquired by a built-in multi-sensor of a mobile terminal, and the detected attribute is used for enriching the basic data of the pedestrian network. The pedestrian navigation system brings the road network attribute into the measurement factor of the path planning during the path planning process, provides the optimal path meeting the individual special requirements instead of the shortest path, improves the use feeling of the pedestrian to the navigation system, and provides convenience for the use of users.
Exemplary device
Referring to fig. 4, a device for detecting attributes of a pedestrian network based on a mobile terminal in an embodiment of the present invention is shown, including:
the acquisition module 41 is used for acquiring sensor data by mobile terminals of a plurality of users, and extracting required sensor data after acquiring original data; recording road network attributes of data collected by part of user mobile terminals, and using the road network attributes to construct training set data, and directly uploading data collected by the rest of user mobile terminals to a cloud end to obtain crowdsourcing data, and using the crowdsourcing data as track data for attribute detection;
the construction module 42 is configured to preprocess data used for constructing a training set, calculate a feature value of each sample and label a corresponding attribute type according to sliding window sampling, screen the feature value by using a feature selection algorithm in machine learning, and construct a training data set; as particularly described above;
the classification module 43 is configured to perform classification training on sample data with different attributes by using the obtained training data set and using a machine learning method, so as to obtain a classification model suitable for road network attributes; as particularly described above;
the fusion module 44 is configured to perform data processing on the crowd-sourced trajectory data, and perform attribute detection by using a trained classification model; performing data fusion on the position information and the attribute information of the track data to obtain GPS data with attribute information; as particularly described above;
the matching module 45 is used for performing map matching based on basic data of the existing pedestrian road network and endowing the attribute information of the track points to the position points matched to the existing road network; obtaining a unique label of the position point by using a majority voting method for the position point matched with the plurality of attribute labels; and correcting the abnormal label to obtain the road network data with the attribute information.
Referring to fig. 5, the mobile terminal based pedestrian network attribute detection apparatus 1800 of the present invention may comprise one or more of the following components: processing component 1802, memory 1804, power component 1806, multimedia component 1806, audio component 1810, input/output (I/O) interface 1812, sensor component 1814, and communications component 1816.
The processing component 1802 generally controls the overall operation of the device 1800, such as operations associated with display, telephone calls, data communications, camera operations, and recording operations. The processing component 1802 may include one or more processors 1820 to execute instructions to perform all or part of the steps of the methods described above. Further, the processing component 1802 may include one or more modules that facilitate interaction between the processing component 1802 and other components. For example, the processing component 1802 can include a multimedia module to facilitate interaction between the multimedia component 1806 and the processing component 1802.
The memory 1804 is configured to store various types of data to support operation at the device 1800. Examples of such data include instructions for any application or method operating on the device 1800, contact data, phonebook data, messages, pictures, videos, and so forth. The memory 1804 may be implemented by any type or combination of volatile or non-volatile memory devices, such as Static Random Access Memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic or optical disks.
The power supply component 1806 provides power to the various components of the device 1800. The power components 1806 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power for the apparatus 1800.
The multimedia component 1806 includes a screen providing an output interface between the apparatus 1800 and a user. In some embodiments, the screen may include a Liquid Crystal Display (LCD) and a Touch Panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive an input signal from a user. The touch panel includes one or more touch sensors to sense touch, slide, and gestures on the touch panel. The touch sensor may not only sense the boundary of a touch or slide action, but also detect the duration and pressure associated with the touch or slide operation. In some embodiments, the multimedia component 1806 includes a front facing camera and/or a rear facing camera. The front-facing camera and/or the back-facing camera may receive external multimedia data when the device 1800 is in an operational mode, such as a shooting mode or a video mode. Each front camera and rear camera may be a fixed optical lens system or have a focal length and optical zoom capability.
I/O interface 1812 provides an interface between processing component 1802 and peripheral interface modules, which may be keyboards, click wheels, buttons, etc. These buttons may include, but are not limited to: a home button, a volume button, a start button, and a lock button.
The sensor component 1814 includes one or more sensors for providing various aspects of state assessment for the apparatus 1800. For example, the sensor assembly 1814 can detect the open/closed state of the device 1800, the relative positioning of components, such as the display and keypad of the apparatus 1800, the sensor assembly 1814 can also detect a change in the position of the apparatus 1800 or a component of the apparatus 1800, the presence or absence of user contact with the apparatus 1800, orientation or acceleration/deceleration of the apparatus 1800, and a change in the temperature of the apparatus 1800. Sensor assembly 1814 may include a proximity sensor configured to detect the presence of a nearby object without any physical contact. The sensor assembly 1814 may also include a light sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, the sensor assembly 1814 may also include an acceleration sensor, a gyroscope sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.
The communication component 1816 is configured to facilitate communications between the apparatus 1800 and other devices in a wired or wireless manner. The device 1800 may access a wireless network based on a communication standard, such as WiFi, 2G or 3G, or a combination thereof. In an exemplary embodiment, the communication section 1816 receives a broadcast signal or broadcast associated information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component 1816 further includes a Near Field Communication (NFC) module to facilitate short-range communications. For example, the NFC module may be implemented based on Radio Frequency Identification (RFID) technology, infrared data association (IrDA) technology, Ultra Wideband (UWB) technology, Bluetooth (BT) technology, and other technologies.
In an exemplary embodiment, the apparatus 1800 may be implemented by one or more Application Specific Integrated Circuits (ASICs), Digital Signal Processors (DSPs), Digital Signal Processing Devices (DSPDs), Programmable Logic Devices (PLDs), Field Programmable Gate Arrays (FPGAs), controllers, micro-controllers, microprocessors or other electronic components for performing the above-described methods.
The embodiment of the invention provides equipment. The apparatus includes a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the one or more programs including instructions for:
s210: the method comprises the steps that sensor data are collected through a mobile terminal of a user, road network attributes of the data collected by a part of user mobile terminals are recorded and used for constructing training set data, and the data collected by the rest of user mobile terminals are directly uploaded to a cloud to obtain crowdsourcing data and used as track data for attribute detection.
In the embodiment of the invention, a plurality of user smart phones can be adopted to collect data of a plurality of sensors and extract data of three sensors, namely an accelerometer, a barometer and a GPS; and obtaining triaxial acceleration, an air pressure value and GPS data, and reserving timestamp data.
S220: preprocessing data used for constructing a training set, sampling according to a sliding window, calculating a characteristic value of each sample, marking a corresponding attribute type, screening the characteristic values by using a characteristic selection algorithm in machine learning, and constructing the training data set;
in the embodiment of the present invention, step S220 includes:
s221, noise data in original data used for constructing a training set are removed, the air pressure value cannot be directly used for representing real air pressure change due to obvious vibration of a mobile phone in a motion state, filtering processing needs to be carried out on the air pressure value, and then the original data used for constructing the training set are obtained;
s222, performing data sampling on the obtained data for constructing the training set through a sliding window to obtain a plurality of samples;
s223, after data sampling is completed, selecting a mean value, a variance, a correlation coefficient and a gas pressure difference as initial characteristics of the samples, calculating characteristic values of each sample, reserving a first piece of time stamp data of each sample, screening the initial characteristics by adopting a characteristic selection function of Weka, and extracting an optimal characteristic subset;
and S224, after the characteristic values of the samples are obtained, adding the corresponding attribute labels to obtain the required training data set.
The step of eliminating the noise data in the collected data comprises the following steps:
determining a sampling frequency according to the acquired data; setting a variance threshold value and a rejection amount through sampling frequency, wherein the threshold value is used for judging whether the data are collected in a motion state; the elimination quantity is used for deleting the initial and final data; after the variance threshold value is determined, the variance of the acceleration of the collected data in the z-axis direction is calculated.
S230: according to the obtained training data set, carrying out classification training on sample data with different attributes by adopting a machine learning method to obtain a classification model;
wherein, the step S230 specifically includes:
s231, the obtained final training set is used as model input, and a K-adjacent model is used as a classification model for training to obtain the classification model suitable for the road network attribute.
S240: carrying out attribute detection on the preprocessed crowdsourced track data by using a trained classification model, and carrying out data fusion on attribute information and position information of the track data to obtain GPS data with attribute information;
the step S240 includes:
s241, in the same preprocessing mode as the training data, eliminating noise data in crowdsourced trajectory data, filtering air pressure values, then sampling data, calculating characteristic values of all samples according to selected characteristics in the training data, and meanwhile keeping first time stamp data of all samples;
s242, using the trained KNN model to perform attribute detection on the processed track data, so as to obtain attribute information of each sample of the track data;
s243, after obtaining the attribute information with time information, carrying out data fusion on the attribute information and the position information;
and by comparing the time stamps and the length of the sampling window, the attribute label of each sample is given to the track segment data corresponding to the sample, so that the data fusion of the attribute information and the position information is realized, and the GPS data of each track can be provided with the attribute information.
S250: map matching is performed based on basic data of the existing pedestrian road network, and attribute information of track points is given to position points matched to the existing road network. And for the position points matched with the plurality of attribute labels, obtaining the unique label of the position point by using a majority voting method. And after the abnormal label is corrected, the road network data with the attribute information is obtained.
Wherein the step S250 includes:
and S251, acquiring basic data of the existing road network, sampling the fused track data according to a certain time window length, and performing attribute matching on each sample as a track segment according to the position. The actual road network is a connection combination of a plurality of line segments, and each line segment is used as a matched candidate line segment. And sequentially calculating the distance from each GPS point in the track section to the candidate road section through the Euclidean distance, and summing to express the distance index from the track section to the road section. And after the distance indexes from each sample track section to each candidate road section are obtained, selecting the line section with the closest distance as a matching object. After the road network sections matched with the track sections are determined, because the position points in the road network are the characteristic points of the sections, each section only has two starting points and two ending points, the number of points between the two position points of the road network section is required to be increased in a difference mode, and the accuracy of position matching is improved. According to the Euclidean distance, obtaining points of the track segment, which are closest to the matched object, and assigning the attributes of the track points to the matched position points;
s252, after matching is completed, because the road network point density after point number increase and the track point density are still different, the position points of a part of road network are matched to a plurality of attribute tags, and a plurality of attribute tags of each position point are voted by using a majority voting method to obtain a unique tag;
and S253, due to the problem of classification accuracy, one or a few abnormal attributes may appear in the continuous track points with the same attribute, and the abnormal labels need to be logically judged and corrected, so that the accuracy of road network attribute detection is improved.
Embodiments of the present invention also provide a non-transitory computer readable storage medium, such as the memory 1804, including instructions that are executable by the processor 1820 of the apparatus 1800 to perform the above-described method. For example, the non-transitory computer readable storage medium may be a ROM, a Random Access Memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, and the like.
A non-transitory computer readable storage medium, wherein instructions of the storage medium, when executed by a processor of an electronic device, enable the electronic device to execute the method for detecting a pedestrian network attribute based on a mobile terminal according to the above embodiments, as described above.
Compared with the prior art, the embodiment of the invention has the following advantages:
according to the method provided by the embodiment of the invention, the mobile terminal is used for acquiring data of a plurality of sensors, processing the data, labeling attribute labels, and selecting appropriate sample characteristics for training a classification model so as to realize the function of detecting the attributes of the track data. Road attributes in the pedestrian network are automatically detected and added into basic data of the pedestrian network, the current situation that attribute factors are not considered in path planning is improved, and a data basis is provided for realizing planning of an optimal path and meeting special requirements of pedestrian traveling.
According to the invention, the road attribute of the pedestrian network is detected by adopting the data acquired by the built-in sensor of the mobile terminal, so that after the basic data of the pedestrian network is enriched, the personalized navigation service meeting specific requirements can be provided, and the convenient and comfortable navigation service is provided for travel groups.
Other embodiments of the invention will be apparent to those skilled in the art from consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention following, in general, the principles of the invention and including such departures from the present disclosure as come within known or customary practice within the art to which the invention pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the invention being indicated by the following claims.
It will be understood that the invention is not limited to the precise arrangements described above and shown in the drawings and that various modifications and changes may be made without departing from the scope thereof. The scope of the invention is only limited by the appended claims
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.
Claims (10)
1. A method for detecting attributes of a pedestrian network based on a mobile terminal is characterized by comprising the following steps:
the method comprises the steps that the mobile terminals of a plurality of users collect sensor data, and required sensor data are extracted after the original data are obtained; recording road network attributes of data collected by part of user mobile terminals, and using the road network attributes to construct training set data, and directly uploading data collected by the rest of user mobile terminals to a cloud end to obtain crowdsourcing data, and using the crowdsourcing data as track data for attribute detection;
preprocessing data used for constructing a training set, sampling according to a sliding window, calculating a characteristic value of each sample, marking a corresponding attribute type, screening the characteristic values by using a characteristic selection algorithm in machine learning, and constructing the training data set;
according to the obtained training data set, carrying out classification training on sample data with different attributes by adopting a machine learning method to obtain a classification model suitable for the road network attributes;
carrying out data processing on the crowdsourced trajectory data, and carrying out attribute detection by using a trained classification model; performing data fusion on the position information and the attribute information of the track data to obtain GPS data with attribute information;
map matching is carried out based on basic data of the existing pedestrian road network, and attribute information of track points is given to position points matched to the existing road network; obtaining a unique label of the position point by using a majority voting method for the position point matched with the plurality of attribute labels; and correcting the abnormal label to obtain the road network data with the attribute information.
2. The method according to claim 1, wherein the mobile terminals of the users collect sensor data, and extract the required sensor data after obtaining raw data; the method comprises the following steps of recording road network attributes of data collected by part of user mobile terminals, constructing training set data, directly uploading data collected by the rest of user mobile terminals to a cloud end to obtain crowdsourcing data, and using the crowdsourcing data as track data for attribute detection:
the method comprises the steps that data of a plurality of sensors are collected through mobile terminals of a plurality of users, and data of an accelerometer, a barometer and a GPS are extracted; and obtaining triaxial acceleration, an air pressure value and GPS data, and reserving timestamp data.
3. The method according to claim 1, wherein the step of preprocessing the data for constructing the training set, calculating the feature value of each sample and labeling the corresponding attribute type according to sliding window sampling, and screening the feature values by using a feature selection algorithm in machine learning comprises:
eliminating noise data in original data used for constructing a training set, and filtering an air pressure value to obtain initial data used for constructing the training set;
carrying out data sampling on the obtained data for constructing the training set through a sliding window to obtain a plurality of samples of the data set;
after data sampling is finished, selecting a mean value, a variance, a correlation coefficient and a gas pressure difference as initial characteristics of samples, calculating a characteristic value of each sample, retaining first time stamp data of each sample, screening the initial characteristics by adopting a characteristic selection function of Weka, and extracting an optimal characteristic subset;
and after the characteristic value of each sample is obtained, adding the corresponding attribute label to obtain the finally required training set data.
4. The method according to claim 3, wherein said step of removing noise data in the data used for constructing the training set comprises:
determining a sampling frequency according to the acquired data;
setting a variance threshold value and a rejection amount according to the calculated sampling frequency;
removing the initial and final data according to the removal amount;
and eliminating the data collected in the non-motion state according to the set variance threshold.
5. The method according to claim 1, wherein the step of performing classification training on sample data with different attributes by using a machine learning method according to the obtained training data set to obtain a classification model suitable for road network attributes comprises:
and taking the obtained final training set data as model input, and training by taking a K-adjacent model as a classification model to obtain the classification model suitable for the road network attribute.
6. The method according to claim 1, wherein the crowd-sourced trajectory data is subjected to data processing, and the trained classification model is used for performing attribute detection; the step of performing data fusion on the position information and the attribute information of the track data to obtain the GPS data with the attribute information comprises the following steps:
preprocessing and sampling crowdsourced trajectory data, calculating characteristic values of all samples according to selected characteristics in training data, and meanwhile, reserving first time stamp data of each sample;
performing attribute detection on the processed track data by using the trained K-adjacent model to obtain attribute information of each sample of the track data and time information;
and performing data fusion on the obtained attribute information with the time information and the position information to obtain the GPS data with the attribute information of each track.
7. The method for detecting the attribute of the pedestrian network based on the mobile terminal according to claim 1, wherein map matching is performed based on the basic data of the existing pedestrian network, and attribute information of the track point is given to the position point matched to the existing pedestrian network; obtaining a unique label of the position point by using a majority voting method for the position point matched with the plurality of attribute labels; the step of correcting the abnormal label to obtain the road network data with the attribute information comprises the following steps:
acquiring basic data of the existing road network, sampling the fused track data according to a certain time window length, and performing attribute matching on each sample as a track segment according to the position;
after matching is finished, due to the difference between the road network point density and the track point density, the position points of a part of road network are matched to a plurality of attribute tags, and a plurality of attribute tags of each position point are voted by using a majority voting method to obtain a unique tag.
After the unique label is obtained, due to the influence of classification precision, the attribute detected by part of position points is wrong, logic judgment is needed, and an abnormal label is obtained and corrected to obtain the final road network basic data with the attribute.
8. A pedestrian network attribute detection device based on a mobile terminal is characterized by comprising:
the acquisition module is used for acquiring sensor data by mobile terminals of a plurality of users, and extracting required sensor data after acquiring original data; recording road network attributes of data collected by part of user mobile terminals, and using the road network attributes to construct training set data, and directly uploading data collected by the rest of user mobile terminals to a cloud end to obtain crowdsourcing data, and using the crowdsourcing data as track data for attribute detection;
the construction module is used for preprocessing data used for constructing a training set, calculating a characteristic value of each sample according to sliding window sampling, marking a corresponding attribute type, screening the characteristic values by using a characteristic selection algorithm in machine learning, and constructing the training data set;
the classification module is used for performing classification training on sample data with different attributes by using the obtained training data set and adopting a machine learning method to obtain a classification model suitable for road network attributes;
the fusion module is used for carrying out data processing on the crowdsourced trajectory data and carrying out attribute detection by using a trained classification model; performing data fusion on the position information and the attribute information of the track data to obtain GPS data with attribute information;
the matching module is used for carrying out map matching based on basic data of the existing pedestrian road network and endowing the attribute information of the track points to the position points matched to the existing road network; obtaining a unique label of the position point by using a majority voting method for the position point matched with the plurality of attribute labels; and correcting the abnormal label to obtain the road network data with the attribute information.
9. A computer device comprising a memory, and one or more programs, wherein the one or more programs are stored in the memory, and configured to be executed by one or more processors comprises the steps for performing the mobile terminal based pedestrian network property detection method according to any of claims 1-7.
10. A non-transitory computer readable storage medium, wherein instructions, when executed by a processor of an electronic device, enable the electronic device to perform the steps of the mobile terminal based pedestrian network property detection method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010104216.8A CN111288999B (en) | 2020-02-19 | 2020-02-19 | Pedestrian road network attribute detection method, device and equipment based on mobile terminal |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010104216.8A CN111288999B (en) | 2020-02-19 | 2020-02-19 | Pedestrian road network attribute detection method, device and equipment based on mobile terminal |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111288999A true CN111288999A (en) | 2020-06-16 |
| CN111288999B CN111288999B (en) | 2021-08-31 |
Family
ID=71025556
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010104216.8A Active CN111288999B (en) | 2020-02-19 | 2020-02-19 | Pedestrian road network attribute detection method, device and equipment based on mobile terminal |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111288999B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112013862A (en) * | 2020-07-31 | 2020-12-01 | 深圳大学 | Pedestrian network extraction and updating method based on crowdsourcing trajectory |
| CN112380316A (en) * | 2020-12-09 | 2021-02-19 | 浙江浙蕨科技有限公司 | Travel situation data processing method and storage medium |
| CN112527930A (en) * | 2020-11-19 | 2021-03-19 | 深圳大学 | Pedestrian road gradient information extraction method based on smart phone |
| CN113848878A (en) * | 2021-08-24 | 2021-12-28 | 深圳大学 | Indoor and outdoor three-dimensional pedestrian road network construction method based on multi-source data |
| CN114359774A (en) * | 2021-11-17 | 2022-04-15 | 山东省国土测绘院 | Pedestrian movement mode classification method and device and electronic equipment |
| CN115081505A (en) * | 2022-01-18 | 2022-09-20 | 中国地质大学（武汉） | Pedestrian network incremental generation method based on walking track data |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105263113A (en) * | 2015-10-26 | 2016-01-20 | 深圳大学 | Wi-Fi location fingerprint map building method and system based on crowd-sourcing |
| CN105426826A (en) * | 2015-11-09 | 2016-03-23 | 张静 | Tag noise correction based crowd-sourced tagging data quality improvement method |
| CN105957342A (en) * | 2016-05-30 | 2016-09-21 | 武汉大学 | Lane-level road mapping method and system based on crowdsourcing space-time big data |
| CN107330396A (en) * | 2017-06-28 | 2017-11-07 | 华中科技大学 | A kind of pedestrian's recognition methods again based on many attributes and many strategy fusion study |
| CN107688665A (en) * | 2017-09-27 | 2018-02-13 | 深圳大学 | A kind of indoor map method for auto constructing, device and storage medium |
| CN109215347A (en) * | 2018-10-22 | 2019-01-15 | 北京航空航天大学 | A kind of traffic data quality control method based on crowdsourcing track data |
| CN109816982A (en) * | 2019-02-25 | 2019-05-28 | 东南大学 | Virtual networks non-motorized lane attribute modification method based on shared bicycle track |
| JP2019164761A (en) * | 2018-01-29 | 2019-09-26 | 楽天株式会社 | Improving quality of labeled training data |
| CN110287276A (en) * | 2019-05-27 | 2019-09-27 | 百度在线网络技术（北京）有限公司 | High-precision map updating method, device and storage medium |
| US20190355173A1 (en) * | 2018-05-16 | 2019-11-21 | Samsung Electronics Co., Ltd. | Leveraging crowdsourced data for localization and mapping within an environment |
-
2020
- 2020-02-19 CN CN202010104216.8A patent/CN111288999B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105263113A (en) * | 2015-10-26 | 2016-01-20 | 深圳大学 | Wi-Fi location fingerprint map building method and system based on crowd-sourcing |
| CN105426826A (en) * | 2015-11-09 | 2016-03-23 | 张静 | Tag noise correction based crowd-sourced tagging data quality improvement method |
| CN105957342A (en) * | 2016-05-30 | 2016-09-21 | 武汉大学 | Lane-level road mapping method and system based on crowdsourcing space-time big data |
| CN107330396A (en) * | 2017-06-28 | 2017-11-07 | 华中科技大学 | A kind of pedestrian's recognition methods again based on many attributes and many strategy fusion study |
| CN107688665A (en) * | 2017-09-27 | 2018-02-13 | 深圳大学 | A kind of indoor map method for auto constructing, device and storage medium |
| JP2019164761A (en) * | 2018-01-29 | 2019-09-26 | 楽天株式会社 | Improving quality of labeled training data |
| US20190355173A1 (en) * | 2018-05-16 | 2019-11-21 | Samsung Electronics Co., Ltd. | Leveraging crowdsourced data for localization and mapping within an environment |
| CN109215347A (en) * | 2018-10-22 | 2019-01-15 | 北京航空航天大学 | A kind of traffic data quality control method based on crowdsourcing track data |
| CN109816982A (en) * | 2019-02-25 | 2019-05-28 | 东南大学 | Virtual networks non-motorized lane attribute modification method based on shared bicycle track |
| CN110287276A (en) * | 2019-05-27 | 2019-09-27 | 百度在线网络技术（北京）有限公司 | High-precision map updating method, device and storage medium |
Non-Patent Citations (3)
| Title |
|---|
| XUE YANG 等: "Pedestrian network generation based on crowdsourced tracking data", 《INTERNATIONAL JOURNAL OF GEOGRPHICAL INFORMATION SCIENCE》 * |
| 吴迪: "基于多轨迹GPS数据和路网属性的个性化行人导航系统设计与实现", 《中国优秀硕士学位论文全文数据库 工程科技II辑》 * |
| 谢金运等: "大规模浮动车流数据并行地图匹配方法", 《武汉大学学报(信息科学版)》 * |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112013862A (en) * | 2020-07-31 | 2020-12-01 | 深圳大学 | Pedestrian network extraction and updating method based on crowdsourcing trajectory |
| CN112013862B (en) * | 2020-07-31 | 2022-06-17 | 深圳大学 | Pedestrian network extraction and updating method based on crowdsourcing trajectory |
| CN112527930A (en) * | 2020-11-19 | 2021-03-19 | 深圳大学 | Pedestrian road gradient information extraction method based on smart phone |
| CN112527930B (en) * | 2020-11-19 | 2023-11-03 | 深圳大学 | Pedestrian road network road gradient information extraction method based on smart phone |
| CN112380316A (en) * | 2020-12-09 | 2021-02-19 | 浙江浙蕨科技有限公司 | Travel situation data processing method and storage medium |
| CN113848878A (en) * | 2021-08-24 | 2021-12-28 | 深圳大学 | Indoor and outdoor three-dimensional pedestrian road network construction method based on multi-source data |
| CN113848878B (en) * | 2021-08-24 | 2023-08-08 | 深圳大学 | Indoor and outdoor three-dimensional pedestrian road network construction method based on crowd source data |
| CN114359774A (en) * | 2021-11-17 | 2022-04-15 | 山东省国土测绘院 | Pedestrian movement mode classification method and device and electronic equipment |
| CN114359774B (en) * | 2021-11-17 | 2023-04-07 | 山东省国土测绘院 | Pedestrian movement mode classification method and device and electronic equipment |
| CN115081505A (en) * | 2022-01-18 | 2022-09-20 | 中国地质大学（武汉） | Pedestrian network incremental generation method based on walking track data |
| CN115081505B (en) * | 2022-01-18 | 2024-10-15 | 中国地质大学（武汉） | Pedestrian road network incremental generation method based on walking track data |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111288999B (en) | 2021-08-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111288999B (en) | Pedestrian road network attribute detection method, device and equipment based on mobile terminal | |
| US9763055B2 (en) | Travel and activity capturing | |
| CN108827307B (en) | Navigation method, navigation device, terminal and computer readable storage medium | |
| CN108197327B (en) | Song recommendation method, device and storage medium | |
| CN103914559A (en) | Network user screening method and network user screening device | |
| CN104537837B (en) | Method, device and system for determining traffic routes | |
| CN112749825B (en) | Method and device for predicting destination of vehicle | |
| CN104969033A (en) | Method and apparatus for navigation based on media density along possible routes | |
| US20150186426A1 (en) | Searching information using smart glasses | |
| CN106228158A (en) | The method and apparatus of picture detection | |
| CN112556717B (en) | Travel mode screening method and travel route recommending method and device | |
| CN107391604A (en) | Map styles display methods, device and mobile terminal based on active state | |
| CN106203306A (en) | The Forecasting Methodology at age, device and terminal | |
| CN107527024A (en) | Face face value appraisal procedure and device | |
| CN110929176A (en) | Information recommendation method and device and electronic equipment | |
| CN112115894A (en) | Training method and device for hand key point detection model and electronic equipment | |
| CN107341226B (en) | Information display method and device and mobile terminal | |
| CN107786746A (en) | Control method, device and system, the storage medium of traffic application | |
| CN108241678B (en) | Method and device for mining point of interest data | |
| CN106323296A (en) | Method and apparatus for recognizing travel model | |
| US20240328803A1 (en) | Methods and Systems for Determining Estimated Travel Times Through a Navigable Network | |
| CN111797655A (en) | User activity identification method and device, storage medium and electronic equipment | |
| KR102188008B1 (en) | Electronic device, method and system for providing contents relate to traffic | |
| CN114550115A (en) | Target tracking method, device, terminal and storage medium | |
| CN113077370A (en) | City function attribute identification method, equipment, medium and product of geographic tile |
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 |