JP5520203B2 - Congestion degree estimation device - Google Patents

Description

  The present invention relates to a congestion degree estimation device.

  As a background art of this technical field, Patent Document 1 describes a congestion estimation system, which includes a motion information generation unit that calculates motion information from an image sequence, and “frequency transform processing using Fourier transform on an image. Go ”texture information generation unit that generates texture information of the image, motion information determination unit that determines whether there is motion, texture information determination unit that determines whether there is texture information that is the same as a person, motion information And a stay determination unit that determines whether or not a person exists based on the result of texture information determination, and uses the presence / absence of motion and similarity determination using texture to determine various states for each region. There is disclosed a technique for discriminating and estimating the degree of congestion and providing information on whether or not it is an congestion condition and an abnormal state.

  Further, Patent Document 2 describes a congestion information notification system, in which each vehicle gets on the basis of the weight of each vehicle detected by a variable load device that detects the weight of the vehicle provided on the train or each train vehicle. Establishing the rate, transmitting information on the congestion status in each vehicle to the station, and providing a home monitoring device for monitoring the congestion status on the platform, A technique is also disclosed in which both the passenger on the platform and a person outside the station are provided with a large image display device (Aurora Vision) outside the station.

  Patent Document 3 describes a system for collecting and providing train congestion rate information, which monitors the passengers for each train entering and leaving the station, and uses the moving images of the passengers to get on and off passengers. The number of passengers remaining in the train for each train is calculated from the number of passengers and the number of passengers, and the congestion rate for each train is calculated from the number of passengers. A system for collecting and providing the congestion rate is disclosed.

  Patent documents 4 and 5 describe a train operation management system, an operation management device, and the like. When the diagram management device detects a schedule change such as a train delay or an operation adjustment input, the arrival and departure of the train Calculate the predicted value of the time, output this predicted value to the train operation information providing server, or edit the information from the operation management system and display it on the on-line status monitor screen. A technique is disclosed in which, when a body is designated, a corresponding ground station is activated and a communication line is formed with a mobile body that is on the line.

  Patent Document 6 discloses a technology of a vehicle detection device that detects whether a train has arrived at a station.

JP 2009-110152 A JP 2000-190847 A JP 2007-1381 A JP 2007-18149 A JP 2002-59836 A JP 2007-45348 A

  Japanese Patent Application Laid-Open No. 2004-228561 describes that image data captured by a camera is processed to estimate the congestion level, and that the congestion level is determined to be abnormal when the estimated congestion level does not fall below a predetermined threshold.

  However, for example, in the case of a station that is normally congested, such as during morning and evening rush hours in a major terminal station in a metropolitan area, even if the congestion degree is estimated using only image data acquired at that time It is very difficult to distinguish whether it is "normal" congestion that is seen on average during normal morning and evening rush hours, or unusual congestion caused by some cause such as an accident.

  In addition, even if it is estimated that unusual congestion occurred, if the train is operating normally according to the schedule, there is no problem in the operation management of the train, and the congestion is independent of the train operation status. If the degree is estimated to be abnormal, there is a risk of hindrance to train operation management.

  Therefore, the present invention has an object to provide a congestion degree estimation device that estimates the degree of congestion such as whether the congestion is normal or abnormal in consideration of current and past train operation conditions and change patterns of human movement. .

The congestion degree estimation apparatus according to the present invention includes an imaging unit that acquires video data near a point where a moving object arrives, a human flow measurement unit that performs image processing on the video data and measures a person's flow, and measurement by the human flow measurement unit A pattern database for storing pattern data representing typical time-series changes of results, a mobile object detection unit for detecting arrival of the mobile object, a detection result of the mobile object obtained by the mobile object detection unit, and A history database creation unit that stores measurement results of the human flow measurement unit in a history database, an analysis unit that extracts pattern data representing typical time-series changes from the measurement results of the human flow measurement unit, and operation of the mobile body comprising a determining pedestrian flow change determination unit congestion degree by referring to the schedule information and the human flow measurement portion of the measurement result and the pattern data, and a display unit that displays the congestion, the analysis Statistically processing the history data stored in the history database, creating pattern data representing typical time series changes for each predetermined classification class, and storing the pattern data in the pattern database, and changing the human flow The determination unit refers to the operation schedule information of the mobile object, the history database, and the pattern database, and the difference between the detection information from the mobile object detection unit and the operation schedule information of the mobile object, the pattern database, and the history The degree of congestion of a person is determined by calculating the similarity of time series data for a plurality of past data that is closest to the estimated arrival time of the moving object stored in the database, and the determination result of the human flow change determination unit determines the degree of congestion. It is characterized in that the display unit displays whether the moving body or the degree of congestion near the arrival place of the moving body is normal or abnormal.

  The congestion degree estimation apparatus of the present invention is further characterized in that a difference between the pattern data and the processing result of the human flow measurement unit is displayed on the display unit as a deviation level.

  The congestion degree estimation apparatus of the present invention includes an imaging unit that acquires video data inside the moving body, instead of an imaging unit that acquires video data near a point where the moving body arrives. It is characterized in that the degree of congestion of a person is estimated using the video data inside.

  The congestion degree estimation apparatus according to the present invention transmits from the mobile unit instead of an imaging unit that acquires video data near a point where the mobile unit arrives and a human flow measurement unit that performs image processing on the video data to detect a human flow. A passenger number measuring unit that measures the number of passengers of the moving body based on the weight measurement result of the moving body, and estimating the degree of congestion of the person using the measurement result of the passenger number measuring unit Features.

  The congestion degree estimation apparatus according to the present invention further includes an imaging unit that acquires video data inside the moving body, or the number of passengers of the moving body based on a weight measurement result of the moving body that is transmitted from the moving body. The number of passengers measuring unit that measures the degree of congestion of the person in the person flow change determining unit, and the degree of congestion of the person estimated using the video data inside the moving body or the number of passengers measuring unit Based on both determination results of the determination result of the degree of congestion of the person estimated using the measurement result, a congestion determination unit that determines the degree of congestion of the person is provided, and according to the determination result of the congestion determination unit, the It is characterized in that the display unit displays whether the moving body or the degree of congestion near the arrival place of the moving body is normal or abnormal.

  ADVANTAGE OF THE INVENTION According to this invention, the congestion degree estimation apparatus which can determine a congestion degree exactly according to the operation condition of a train can be provided.

  Problems, configurations, and effects other than those described above will be clarified by the following description of embodiments.

FIG. 1 is an example of a configuration diagram in the first embodiment of the congestion degree estimation apparatus of the present invention. FIG. 2 is an example of an image acquired by the imaging unit 101. FIG. 3 is a flowchart showing the processing contents of the human flow measurement unit 105. FIG. 4 shows data representing a background image used in the human flow measurement unit 105. FIG. 5 shows a data structure of the history database 110 stored in the human flow data management unit 109. FIG. 6 is a flowchart showing the processing contents of the analysis unit 107. FIG. 7 is a diagram showing the data structure of the pattern database 111 stored in the human flow data management unit 109. FIG. 8 is a flowchart showing another processing content of the analysis unit 107. FIG. 9 is a flowchart showing the processing contents of the human flow change determination unit 108. FIG. 10 is a graph schematically showing a normal human flow change pattern used in the human flow change determination unit 108. FIG. 11 is a graph schematically showing a human flow change pattern at the time of abnormality used in the human flow change determination unit 108. FIG. 12 is a diagram illustrating an example in which the determination result in the human flow change determination unit 108 is displayed on the display screen 120 in the operation management device 112. FIG. 13 is a view showing a display example in the information providing apparatus 114. FIG. 14 is an example of a configuration diagram in the second embodiment of the congestion degree estimation apparatus of the present invention. FIG. 15 is an example of an image acquired by the imaging unit 206. FIG. 16 is a diagram illustrating a data structure of the passenger number history database 210 stored in the passenger number data management unit 209. FIG. 17 is a diagram illustrating a data structure of the passenger number pattern database 211 stored in the passenger number data management unit 209. FIG. 18 is a flowchart showing the processing contents of the congestion determination unit 217.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

  In the present embodiment, an example of a congestion level estimation device that estimates the congestion level will be described. FIG. 1 is an example of a configuration diagram of a congestion degree estimation apparatus 100 according to the first embodiment of the present invention. The congestion degree estimation device 100 includes an imaging unit 101, an image processing unit 102, a vehicle detection unit 103, a human flow information control unit 104, and a human flow data management unit 109.

  The human flow information control unit 104 includes a human flow measurement unit 105, a history data creation unit 106, an analysis unit 107, and a human flow change determination unit 108. The human flow data management unit 109 has a history database 110 and a pattern database 111. Furthermore, the congestion degree estimation device 100 is connected to the operation management device 112.

  Hereinafter, the operation of each device in the congestion degree estimation device 100 will be described in detail. A plurality of imaging units 101 are installed in the vicinity of a station platform, stairs, or escalator. Then, for example, the image data 121 is obtained so that the person can go in and out of the home as shown in FIG.

  The image processing unit 102 performs processing such as digital conversion and error correction on the image data 121 acquired by the imaging unit 101 and stores the data in a temporary storage medium such as a memory. Then, in response to a request from the human flow measuring unit 105, the stored data is read and written.

  The vehicle detection unit 103 is a device that detects whether or not a train has arrived at a station, and can be realized by the technique disclosed in Patent Document 6, for example.

  The human flow measuring unit 105 performs a person counting process on the image data 121 stored in the image processing unit 102. The measurement result is stored in the history data creation unit 106 as the history database 110 in the human flow data management unit 109.

  Hereinafter, processing of the human flow measurement unit 105 and the history data creation unit 106 will be described in detail. FIG. 3 is a flowchart showing processing procedures of the human flow measurement unit 105 and the history data creation unit 106.

  In FIG. 3, the processing procedure includes a person candidate area detection process in step 301, a feature amount extraction process in step 302, a person determination process in step 303, a human flow measurement process in step 304, and a database storage process in step 305.

  The human candidate area detection process in step 301, the feature amount extraction process in step 302, the person determination process in step 303, and the human flow measurement process in step 304 are processed in the human flow measurement unit 105, and the database storage process in step 305 is history data creation. Processed by the unit 106.

  The person candidate area detection process in step 301 is performed as follows, for example. An image when no person is present in the imaging range as shown in FIG. 4 is stored as a background image 401 in advance, and the difference between the input image 201 and the background image is performed. Thus, the person candidate area is detected.

  The feature amount extraction process in step 302 is performed as follows, for example. A gray edge is detected by using a Sobel filter or a Laplace filter for the input person candidate region. Then, a circle or an ellipse region is detected by performing a Hough transform on the detected edge information.

  The person determination process in step 303 is performed, for example, by determining an area having a size equal to or larger than a predetermined threshold as a head area with respect to the circle or ellipse area obtained by the feature amount extraction process in step 302. .

  The human flow measurement process in step 304 is performed as follows, for example. In the input image 201, a detection area 122 is set. This detection area 122 is set for each moving direction to be discriminated. For example, in the input image in FIG. 2, two directions of up and down are considered as the moving direction of the person. For this reason, two detection areas in the upstream direction and the downstream direction are set as detection areas. If the circle or ellipse process determined to be a person in the person determination process in step 303 exists in the detection area 122, it is determined that the person has passed, and the passage of the person is counted to measure the human flow.

  The database storage process in step 305 operates as follows. The number of passing people counted in the human flow measurement process in step 304 is totaled for each predetermined time unit. Then, the total result data is stored together with the vehicle detection result input from the vehicle detection unit 103 as the history database 110.

  FIG. 5 is a diagram showing an example of the data structure of the history database 110 stored in the human flow data management unit 109. The history database 110 includes elements of a record ID 501, a home ID 502, a camera ID 503, a direction 504, a date and time 505, a measurement number 506, and a vehicle detection attribute 507.

  The record ID 501 is an ID uniquely assigned to each record. The home ID 502 is an ID for identifying which home in the station is installed. The camera ID 503 is an ID for uniquely identifying the installed camera. The direction 504 is information for identifying in which detection area the person detected by the camera is detected. The date and time 505 is the time when the number of people is measured, and is recorded in a predetermined time unit.

  The number of people 506 is the number of people measured at the date and time 505. The vehicle detection attribute 507 is information obtained from the vehicle detection result input from the vehicle detection unit 103. In this embodiment, the stop state in the vehicle station and the waiting state for entering the vehicle are recorded. The history database 110 is stored for a predetermined period.

  FIG. 6 is a flowchart illustrating an example of processing contents of the analysis unit 107. The analysis unit 107 refers to the history database 110 to calculate an average value and variance, and creates a pattern database 111.

  Hereinafter, each process will be described in detail. In step 601, a classification class for specifying a record from which the average value and variance of the number of people to be measured are calculated is determined from the history database 110 stored in the image processing result storage database 106.

  This classification class determines a class so that the change pattern of the day considered to be similar in the number change pattern belongs to the same class as much as possible. In this embodiment, the classification classes are weekdays, days before holidays, and holidays. The average and variance of the number of people passing through are determined for each classification class.

  In step 602, a record corresponding to the classification class determined in step 601 is acquired from the history database 500 based on the date and time 505.

  In step 603, it is determined whether the number of records belonging to the classification class is greater than or equal to a predetermined threshold. If it exists above the threshold, the process proceeds to step 604. If it is less than or equal to the threshold value, the process returns to step 602.

  In step 604, the average value of the number of passing people is calculated from the records in the history database 110 belonging to the corresponding classification class.

  In step 605, the variance value of the number of passing people is calculated from the records in the history database 110 belonging to the corresponding classification class.

  In step 606, vehicle detection attributes to be stored in the statistical database are determined. Specifically, among the vehicle detection attributes 506 existing in the records in the history database belonging to the corresponding classification class, the vehicle detection attribute having the largest number of records is set as the vehicle detection attribute stored in the statistical database.

  In step 607, the pattern database 111 is updated based on the calculated average value, variance, and vehicle detection attribute.

  FIG. 7 is a diagram showing the data structure of the pattern database 111 stored in the human flow data management unit 109. The pattern database 111 includes record ID 701, home ID 702, camera ID 703, direction 704, classification class 705, time 706, vehicle detection attribute 707, average number of people 708, variance 709, and population parameter 710.

  The record ID 701, home ID 702, camera ID 703, and direction 704 are data having the same meaning as the record ID 501, home ID 502, camera ID 503, and direction 504 stored in the history database 500.

  The classification class 705 is the classification class determined in step 601. Time 706 is a reference time for calculating the average value and the variance value, and is obtained by dividing 24 hours at a predetermined interval. The vehicle detection attribute 707, the average number of people 708, and the variance 709 are values of the vehicle detection attribute, the average number of people, and the variance calculated by the analysis unit 107, respectively.

  The parameter 710 is the number of records in the history database 110 used when calculating the average number of people 708 and the variance 709. Or the analysis part 107 may perform the process which extracts the data of a certain day and registers it as a pattern database instead of performing the statistical process which was demonstrated in the said Example.

  FIG. 8 is a flowchart showing another example of processing contents in the analysis unit 107. In step 801, a classification class is determined. This processing is equivalent to the processing content in step 601.

  In step 802, a record for one day in which no abnormal data exists is acquired from the history database 500 for each of the home ID 502, camera ID 503, and direction 504. One way to select a record for a day is to manually select the day on which the train was normally operated.

  In step 803, the record selected in step 802 is registered in the pattern database 111.

  The pattern database created by the process shown in FIG. 8 has a structure in which the variance 709 and the parameter 710 are excluded from the pattern database 111 shown in FIG.

  FIG. 9 is a flowchart showing the processing contents of the human flow change determination unit 108. The human flow change determination unit 108 refers to the history database 110 and the pattern database 111 stored in the human flow data management unit 109 to determine whether or not the change in the human flow is normal.

  Hereinafter, the processing content of the human flow change determination unit 108 will be described in detail. In step 901, schedule schedule information is referred to. Schedule schedule information is train schedule information that can be acquired from the operation management device 112 or the like, and that the train is scheduled to arrive at the platform at what hour and how many minutes.

  In step 902, the history database 110 stored in the human flow data management unit 109 is referred to. The record to be referred to is determined according to the following method, for example. With reference to the schedule diagram information acquired in step 901, the history database 110 is referred to when the train is scheduled to arrive. From the records stored in the history database 110, the past n records whose information in the date / time column 505 is closest to the estimated train arrival time are extracted.

  In step 903, the pattern database 111 stored in the human flow data management unit 109 is referred to. Specifically, first, it is determined which class of the classification classes of the pattern database 111 the current time corresponds to. For example, if the current time is a weekday, it corresponds to a weekday class, and if it is a holiday, it corresponds to a holiday class.

  Then, out of the records stored in the pattern database 111, records for the past n items whose classification class 705 is the class and whose time 706 is closest to the estimated train arrival time are extracted.

  In step 904, it is determined whether or not the time difference Td of the vehicle detection attribute is within a threshold value. Specifically, the following method is used. Of the records in the history database 110 acquired in Step 902, the vehicle detection attribute 507 stored in the record closest to the estimated train arrival time, and among the records in the pattern database 111 acquired in Step 903, the train arrival The vehicle detection attribute 707 stored in the record closest to the scheduled time is compared.

  Then, it is determined whether or not the attribute values of both vehicle detection attributes are the same and the time difference Td is within a certain value range. If it is determined that the time difference Td is within a certain range, the process proceeds to step 905, where it is determined that the train is on the schedule and matches the past driving pattern, and the pattern of the human flow is determined to be normal. On the other hand, if it is determined that the time difference Td does not exist within a certain range, the process proceeds to the next step 906.

In step 906, the time-series data composed of the record of the history database 110 acquired in step 902 is compared with the time-series data composed of the record of the pattern database 111 acquired in step 903. It is determined whether the similarity D of the time series data is within a threshold value. The similarity D between time series data is calculated based on the following calculation formula, for example.

  Here, xi is the record at time i of the history data acquired at step 902, yi is the record at time i of the pattern data acquired at step 903, and n is the number of time-series data, that is, the number of records. It is determined whether or not the value of D is within a predetermined threshold.

  If the calculated value of D is within the predetermined threshold, the process proceeds to step 905, where it is determined that the pattern matches the past human flow change pattern, and the current human flow pattern is determined to be normal. On the other hand, if the calculated value of D is larger than the predetermined threshold value, the process proceeds to step 907, where it is determined that it does not match the past human flow change pattern, and the current human flow pattern is determined to be abnormal.

  FIG. 10 is a graph schematically showing a human flow change pattern in a normal state. A broken line 1001 is time-series data composed of the record of the history database 110, a broken line 1002 is time-series data composed of the record of the pattern database 111, a line 1003 is a scheduled train arrival time, and a line 1004 is a scheduled train arrival time. The time at which history data with the closest vehicle detection attribute being stopped exists, the time when the line 1005 has pattern data with the vehicle detection attribute closest to the scheduled arrival time at the stop, time difference 1006 is the time difference Td of the vehicle detection attribute, A time difference 1007 is the length n of the time series data, and an area 1008 is an area D representing the degree of similarity between the time series data.

  As shown in FIG. 10, in normal times, the history data and the statistical data are highly similar, and as a result, the region D is also small.

  On the other hand, FIG. 11 shows a graph schematically showing a human flow change pattern at the time of abnormality. In FIG. 11, the description of the components having the same reference numerals as those already described in FIG.

  As shown in FIG. 11, the similarity between the history data and the statistical data is low at the time of abnormality, and as a result, the region D becomes relatively large.

  The determination result determined by the human flow change determination unit 108 is transmitted to the operation management device 112 as needed, and is used for train operation management work by the commander. FIG. 12 shows an example in which the determination result in the human flow change determination unit 108 is displayed on the display screen in the operation management apparatus.

  On the display screen 1201 in the operation management device 112, for example, the status of the train line is displayed as described in Patent Document 5. Under the display 1202 indicating the location of the station on the display screen 1201, for example, an icon 1203 indicating the situation of the human flow at each station can be displayed.

  The data displayed on the icon 1203 is a determination result determined by the human flow change determination unit 108, and is displayed as normal if normal, and abnormal if abnormal. As shown in FIGS. 5 and 7, there are a plurality of camera IDs 503 or 703 on the same route of each station.

  For this reason, the determination result displayed on the icon 1203 can be displayed as abnormal if, for example, one of the plurality of camera IDs is determined as abnormal. Furthermore, as indicated by the icon 1204, the size of the area D1008 shown in FIG. 10 or FIG. 11 may be divided into stages, and the difference from the normal time may be displayed as a level.

  Further, the icon 1203 or 1204 may be selectable. When an icon is selected, it is possible to display a normal human flow change pattern and a current human flow change pattern in a representative camera ID at the selected station shown in FIG. 10 or FIG. 11, for example. Or when it displays as abnormal, the information in camera ID determined to be abnormal may be displayed. The commander who performs the train operation management work can instruct an appropriate train operation using the information displayed on the display screen 1201.

  The determination result 1203 displayed on the display screen 1201 in the operation management device 112 or the level difference 1204 may be displayed on the information providing device 114 via the network 113. Examples of the information providing device 114 include an information providing device installed at the entrance of a station or at a platform, a mobile phone possessed by a railway user, a terminal device possessed by a station employee, or the like.

  A display example in the information providing apparatus 114 is shown in FIG. On the display screen 1301, for a station where abnormal congestion has occurred, for example, a determination result 1302 in text display or an icon 1303 in which the level of the difference in congestion from the normal time is displayed. The display content is equivalent to the determination result 1203 and the level difference 1204 displayed on the display screen 1201 in the operation management device 112.

  In the congestion degree estimation apparatus 100 according to the first embodiment of the present invention, it is possible to accurately determine the congestion degree of the station platform according to the train operation status and more accurately display the congestion degree such as whether the congestion is normal or abnormal.

  In the present embodiment, another device configuration example of the device 100 that performs the congestion degree estimation will be described. FIG. 14 is an example of a configuration diagram illustrating the congestion degree estimation apparatus 200 according to the second embodiment of the present invention.

  In the congestion degree estimation apparatus 200 in FIG. 14, the description of the portions having the same functions as those in the configuration denoted by the same reference numerals as those already described in FIG.

  In the congestion degree estimation apparatus 200 according to the second embodiment, in addition to the configuration of the first embodiment, the communication unit 207, the in-train image processing unit 208, the passenger number data management unit 209, the passenger number information control unit 212, and the congestion determination unit 217. It has become the composition which added.

  Hereinafter, the congestion degree estimation apparatus 200 will be described in detail. Each vehicle 201 of the train is provided with a load receiving device 202 and a transmission terminal 204 that transmits the occupancy rate information from each load applying device 202 to the transmission central station 203 of the train. The train includes the transmission central station 203 and the transmission. A communication unit 205 that transmits the occupancy rate information collected at the central office to the occupancy rate measuring unit 206 is provided.

  The variable load device 202, the transmission central station 203, the transmission terminal 204, and the communication unit 205 operate by a method described in Patent Document 2, for example.

  Furthermore, an imaging unit 206 may be installed in each vehicle of the train. A plurality of imaging units 206 are installed at the top of the vehicle. Then, for example, image data 1501 as shown in FIG. The acquired image data 1501 is transmitted to the congestion degree estimation apparatus 200 via the transmission central station 203, the transmission terminal 204, and the communication unit 205 in the same manner as the boarding rate information.

  The image 1501 transmitted to the congestion degree estimation device 200 is input to the in-train image processing unit 208 via the communication unit 207. In the train image processing unit 208, processing equivalent to that of the image processing unit 102 is performed, and the stored data is read and written in response to a request from the passenger number measuring unit 213.

  The number-of-passengers counting unit 213 performs a number-of-peoples measurement process based on the boarding rate information transmitted from the vehicle 201 or the acquired image data 1501. When boarding rate information is transmitted, it is converted into the number of passengers. When image data is transmitted, the person counting process is performed by a method equivalent to the processing procedure of the human flow measuring unit 105 shown in FIG.

  Therefore, by providing the imaging unit 206 that acquires the video data inside the moving body instead of the imaging unit 101 that acquires the video data near the point where the moving body arrives in FIG. The degree of congestion of a person can be estimated using video data inside the moving body.

  Further, as a congestion degree estimation device, instead of an imaging unit that acquires video data near a point where a moving body arrives and a human flow measurement unit that detects human flow by performing image processing on the video data, the mobile unit transmits the video data. By providing a passenger number measuring unit that measures the number of passengers of the moving body based on the weight measurement result of the moving object, it is possible to estimate the degree of congestion of the person using the measurement result of the passenger number measuring unit.

  In FIG. 14, the measurement result is stored in the history data creation unit 214 as a passenger number history database 210 in the passenger number data management unit 209. The processing content of the history data creation unit 214 is equivalent to the processing of the history data creation unit 106.

  FIG. 16 is a diagram showing an example of the data structure of the passenger number history database 210 stored in the passenger number data management unit 209. The number-of-passengers history database 210 includes elements of a record ID 1601, a train ID 1602, a vehicle ID 1603, a camera ID 1604, a date and time 1605, a number of passengers 1606, and a vehicle detection attribute 1607.

  The record ID 1601 is an ID uniquely assigned to each record. Train ID 1602 is an ID for identifying a train. The vehicle ID 1603 is an ID for identifying the vehicle. The camera ID 504 is an ID for uniquely identifying a camera installed in the train. Date and time 1605 is the time when the number of passengers is measured, and is recorded in predetermined time units. The number of passengers 1606 is the number of passengers measured at date 1605. The vehicle detection attribute 1607 is information obtained from the vehicle detection result input from the vehicle detection unit 103. In this embodiment, the vehicle stop state and the vehicle entry state waiting state are recorded. The passenger number history database 210 is stored for a predetermined period.

  The analysis unit 215 refers to the passenger number history database 210 to calculate an average value and variance, and updates the passenger number pattern database 211. The processing method is the same as the processing of the analysis unit 107 in the first embodiment.

  FIG. 17 is a diagram showing a data structure of the passenger number pattern database 211 stored in the passenger number data management unit 209. The passenger number pattern database 211 includes elements of a record ID 1701, a train ID 1702, a vehicle ID 1703, a camera ID 1704, a classification class 1705, a time 1706, a vehicle detection attribute 1707, an average number of people 1708, a variance 1709, and a population number 1710.

  The record ID 1701, train ID 1702, vehicle ID 1703, and camera ID 1704 are data having the same meaning as the record ID 1501, train ID 1502, vehicle ID 1503, and camera ID 1504 in the passenger number history database 210. The classification class 1705, time 1706, vehicle detection attribute 1707, average number of persons 1708, variance 1709, and population parameter 1710 are elements existing in the pattern database 111, classification class 705, time 706, vehicle detection attribute 707, average number of persons 708. , Variance 709, and parameter 710.

  The passenger number change determination unit 216 refers to a database stored in the passenger number data management unit 209 to determine whether the change in the passenger flow in the train is normal. The operation of the passenger number change determination unit 216 is equivalent to the processing of the person flow change determination unit 108 in the first embodiment.

  The congestion determination unit 217 determines whether or not the change pattern of the human flow is normal based on the processing result in the human flow change determination unit 108 and the processing result in the passenger number change determination unit 216, and transmits the determination result to the operation management device 112. To do.

  FIG. 18 is a flowchart showing the processing contents of the congestion determination unit 217. In step 1801, the determination result in the human flow change determination unit 108 is referred to.

  In step 1802, the determination result in the passenger number change determination unit 216 is referred to. In step 1803, it is determined whether or not the human flow change is normal. If it is determined to be normal, the process proceeds to step 1804. On the other hand, if it is determined that it is not normal, the process proceeds to step 1805.

  In step 1805, it is determined whether or not the change in the number of passengers is normal. If it is determined to be normal, the process proceeds to step 1804 to determine normal congestion. On the other hand, if it is determined that it is not normal, the process proceeds to step 1806, where it is determined that there is abnormal congestion.

  The determination result determined by the congestion determination unit 217 is transmitted to the operation management device 112 as needed, and is used for train operation management work by the commander. The method of use is equivalent to the method shown in the first embodiment.

  In the congestion degree estimation device 200 according to the second embodiment of the present invention, it is possible to accurately determine the congestion degree of both the station platform and the arriving train according to the train operation state, and more accurately determine the congestion degree such as whether the congestion is normal or abnormal. Can be displayed.

  In addition, this invention is not limited to an above-described Example, Various modifications are included. For example, the above-described embodiments have been described in detail for easy understanding of the present invention, and are not necessarily limited to those having all the configurations described.

  Further, a part of the configuration of one embodiment can be replaced with the configuration of another embodiment, and the configuration of another embodiment can be added to the configuration of one embodiment. Further, it is possible to add, delete, and replace other configurations for a part of the configuration of each embodiment.

  Each of the above-described configurations, functions, processing units, processing means, and the like may be realized by hardware by designing a part or all of them with, for example, an integrated circuit. Each of the above-described configurations, functions, and the like may be realized by software by interpreting and executing a program that realizes each function by the processor.

  Information such as programs, tables, and files for realizing each function can be stored in a memory, a hard disk, a recording device such as an SSD (Solid State Drive), or a recording medium such as an IC card, an SD card, or a DVD.

  Further, the control lines and information lines indicate what is considered necessary for the explanation, and not all the control lines and information lines on the product are necessarily shown. Actually, it may be considered that almost all the components are connected to each other.

DESCRIPTION OF SYMBOLS 100 Congestion degree estimation apparatus 101 Image pick-up part 102 Image processing part 103 Vehicle detection part 104 Human flow information control part 105 Human flow measurement part 106 History data creation part 107 Analysis part 108 Human flow change determination part 109 Human flow data management part 110 History database 111 Pattern database 112 Operation management device 113 Network 114 Information providing device 120 Display screen 121 Image data 122 Detection area 201 Train 202 Responsive load device 203 Transmission central station 204 Transmission terminal 205 Communication unit 206 Imaging unit 207 Communication unit 208 In-train image processing unit 209 Passenger data Management unit 210 Passenger number history database 211 Passenger number pattern database 212 Passenger number information control unit 213 Passenger number measurement unit 214 History data creation unit 215 Analysis unit 216 Passenger number change determination unit 217 Congestion determination 401 background image 500 history database 500
501 Record ID 501
502 Home ID 502
503 Camera ID 503
504 direction 504
505 Date and time 505
506 Number of people 506
607 Vehicle detection attribute 507
701 Record ID 701
702 Home ID 702
703 Camera ID 703, 704
704 Camera ID
705 Classification class 705
706 Time 706
707 Vehicle detection attribute 707
708 Average number of people 708
709 dispersion 709
710 Parameter 710
1001 Time series data line 1002 of the history database 1002 Time series data line 1003 of the pattern database 1003 Line of scheduled arrival time 1004 Time line 1005 closest to the expected arrival time Time line 1006 where the pattern data of the stop state exists Time difference Td of the vehicle detection attribute,
1007 Length of time series data n,
1008 Region D representing the degree of similarity between time series data
1201 Display screen 1201 Display 1202 indicating the location of the station
1203 Icon 1203 of a determination result indicating the situation of people flow at each station
1204 Level difference icon 1301 Display screen 1302 Judgment result text display 1303 Level difference icon 1501 Image data 1601 Record ID
1602 Train ID
1603 Vehicle ID
1604 Camera ID
1605 Date 1606 Passenger number 1607 Vehicle detection attribute 1701 Record ID
1702 Train ID
1703 Vehicle ID
1704 Camera ID
1705 Classification class 1706 Time 1707 Vehicle detection attribute 1708 Average number of people 1709 Variance 1710 Population

Claims (5)

An imaging unit that acquires video data near the point where the moving object arrives;
A human flow measurement unit that performs image processing on the video data and measures the flow of a person;
A pattern database for storing pattern data representing typical time-series changes in measurement results of the human flow measurement unit;
A moving body detector for detecting the arrival of the moving body;
A history database creation unit that saves the detection result of the moving body obtained by the moving body detection unit and the measurement result of the human flow measurement unit in a history database;
An analysis unit for extracting pattern data representing a typical time series change from the measurement result of the human flow measurement unit;
A flow change determination unit for determining the degree of congestion with reference to the operation schedule information of the moving body, the measurement result of the flow measurement unit, and the pattern data;
A display unit for displaying the degree of congestion,
The analysis unit statistically processes the history data stored in the history database, creates pattern data representing typical time series changes for each predetermined classification class, and stores the pattern data in the pattern database.
The human flow change determination unit refers to the operation schedule information of the mobile object, the history database, and the pattern database, and the difference between the detection information from the mobile object detection unit and the operation schedule information of the mobile object and the pattern database And calculating the similarity of time series data for a plurality of past data closest to the estimated arrival time of the mobile body stored in the history database to determine the degree of congestion of the person,
A congestion degree estimation apparatus, wherein whether or not the moving object or the vicinity of an arrival place of the moving object is normal or abnormal is displayed on the display unit based on a determination result of the human flow change determination unit. In the congestion degree estimation apparatus according to claim 1,
Furthermore, the congestion degree estimation apparatus, wherein a difference between the pattern data and the processing result of the human flow measurement unit is displayed on the display unit as a deviation level. In the congestion degree estimation apparatus according to claim 1,
Instead of an imaging unit that acquires video data in the vicinity of a point where the mobile body arrives, an imaging unit that acquires video data inside the mobile body is provided,
A congestion degree estimation apparatus, wherein the congestion degree of a person is estimated using video data inside the moving body . In the congestion degree estimation apparatus according to claim 1 ,
Weight measurement result of the moving body transmitted from the moving body instead of an imaging section that acquires video data near the point where the moving body arrives and a human flow measurement section that detects the human flow by image processing the video data A passenger number measuring unit for measuring the number of passengers of the moving body based on
Congestion level estimation apparatus characterized that you estimate the degree of congestion of a person using the measurement results of the passengers measuring unit. In the congestion degree estimation apparatus according to claim 1 ,
Further, an imaging unit that acquires video data inside the moving body, or a passenger number measuring unit that measures the number of passengers of the moving body based on a weight measurement result of the moving body transmitted from the moving body,
The person's congestion degree determination result in the human flow change determination unit, and the person's congestion degree estimated using the video data inside the moving body or the person estimated using the passenger number measurement unit measurement result A congestion determination unit that determines the degree of congestion of a person based on both determination results of the determination result of
A congestion degree estimation apparatus , wherein whether or not the moving body or the vicinity of the arrival place of the moving body is normal or abnormal is displayed on the display section based on a determination result of the congestion determination section .

Images