JP5814605B2 - Gate device - Google Patents

Description

  The present invention relates to a gate device, and more particularly to a gate device such as an automatic ticket gate suitable for reliably detecting a weak person with low movement speed such as a wheelchair user or a crutch user and ensuring the safety of passage of the weak person. It is about.

  For example, in conventional traffic control with an automatic ticket gate, control is performed under the same constant conditions regardless of the characteristics of the user. Accordingly, when the user passing through the ticket gate passage of the automatic ticket gate is a wheelchair user, a crutch user, or the like, it takes time to pass through because the traveling speed is slow. In this way, with regard to entrance and exit traffic at gate devices such as automatic ticket gates, those who require time for traffic due to slower traffic speed than normal users (healthy people) are referred to as “weakly-passed persons” below. I will write. Essentially, when a vulnerable person uses an automatic ticket gate, it is desirable to perform traffic control according to the vulnerable person. This is because, based on traffic control under the same conditions, problems such as contact occur between a weak user with different movement speeds and a normal user.

  In order to avoid contact between users in the ticket gate passage of the automatic ticket gate, it is desirable to keep the distance between users as much as possible. For example, by setting the system to delay the receiving operation of the IC card reader (IC card touch unit) and magnetic card reader for a certain time, a certain user always enters the ticket gate passage of the automatic ticket gate. If the certain time has not elapsed, the next user is not accepted and the interval between a certain user and the next user can be increased. However, in such a system setting, the ticket gate capability of the automatic ticket gate is lowered, and the ticket gate processing efficiency is deteriorated. Therefore, it is inherently desirable to reduce the contact between users only when it is determined that the user is a passerby person, and control is performed to increase the interval between subsequent users.

  As related arts for determining whether or not a person is a weak person, there are techniques described in Patent Documents 1 and 2 below.

  The object detection method and the like described in Patent Document 1 can accurately detect vulnerable persons (such as visually impaired persons and wheelchair users when moving at intersections and pedestrian crossings) even from a long distance. This is a method of applying a pulsed laser to a light reflecting object provided on a moving object and returning it as reflected light, based on the distance to the light reflecting object and the reflected light intensity from the light reflecting object. It has a configuration for detecting a moving object. With this configuration, there is a problem that detection cannot be performed unless the moving object is provided with a light reflecting object. In addition, since it is assumed to move at intersections and pedestrian crossings, it has a system that detects at a remote distance systematically, and it is a system that does not match the detection with automatic ticket gates in a relatively short distance range. is there.

The wheelchair-compatible automatic ticket gate described in Patent Document 2 is a mat sensor generated by laying a mat sensor on the floor surface of the ticket gate passage of the automatic ticket gate, and a user moving on the mat sensor when passing through the ticket gate passage. It is made to detect whether it is a wheelchair based on the output signal. However, in this wheelchair-adaptive automatic ticket gate, as a detection method, only a method for comparing the detected sensor signal pattern with a standard template is described. it is about whether or not the way the weak can detect and mention of not, is unknown about the effectiveness of it. In traffic for ticket gates with automatic ticket gates, it is desirable that all vulnerable persons can be accurately detected.

  As a conventional technique related to the present invention, there is a conventional technique described in Patent Document 3 as a gate device configured so that an accurate number of participants can be determined using a pressure sensor.

JP 2006-194617 A Japanese Patent Laid-Open No. 9-212663 JP 2010-204768 A

  As described above, a weak person such as a wheelchair user or a crutch user has a slower moving speed than a normal user, and thus takes an extra time, for example, 2 to 3 seconds in the ticket gate passage of the same distance. Therefore, when a user moves through the ticket gate of an automatic ticket gate at a station, and a movement by a normal user and a movement by a weak person are mixed, a normal user moves following the weak person Then, if a normal user is not aware of the forward traffic weak person, there is a possibility that the users come into contact with each other due to the difference in moving speed. When such a contact state occurs, secondary problems such as a fall accident occur.

  Therefore, in recent traffic control (or entrance / exit management) with gate devices such as automatic ticket gates, whether or not all vulnerable persons (wheelchair users, crutch users, stick staff users, etc.) are vulnerable. It is desired to detect the information quickly with high accuracy and reflect it in the subsequent normal user traffic control.

  In view of the above problems, the object of the present invention is to accurately automatically detect that a passer is a weak passer on the device system side when a normal user and a passerby pass in a mixed manner. To provide a gate device that executes traffic control (entrance / exit management) suitable for traffic, avoids contact and fall accidents between the vulnerable user and normal users, and ensures safe traffic for the vulnerable user is there.

  In order to achieve the above object, the gate device according to the present invention is configured as follows.

The first gate device (corresponding to claim 1) includes a pressure distribution sensor that detects a movement trace of a passerby as pressure distribution data, and a cumulative pressure related to the identification target region based on the pressure distribution data output from the pressure distribution sensor Pressure distribution data management means for generating distribution data and control means for determining the characteristics of a passer-by based on the accumulated pressure distribution data and performing traffic control in accordance with the characteristics.

In the above gate device, for passers who pass through the entrance and exit passages in automatic ticket gates installed at the ticket gates of stations and public security gates, for example, whether or not they are vulnerable Detects passersby's movements as pressure distribution data and generates cumulative pressure distribution data based on this data . When a passer is determined to be a weak passerby, a reading operation for the entry of a subsequent passerby Is stopped by stopping for a certain period of time, thereby opening a space between the vulnerable person and the subsequent normal user to prevent a collision or the like.

The second gate device (corresponding to claim 2) is preferably configured as described above.
An entrance / exit information reader for transmitting / receiving traffic information to / from an entrance / exit medium possessed by a passerby passing through the passageway;
Entry / exit control means for controlling the reading operation of the entry / exit reader ;
Passerby is determined whether the traffic vulnerable based on the cumulative pressure distribution data according to the recognition target area supplied from the pressure distribution data management unit, the passage weak detection signal to the entrance and exit control means when a traffic vulnerable Providing weak traffic detection means,
The entrance / exit control means is characterized by stopping the reading operation of the entrance / exit reading device for a certain period of time when a weak passer detection signal is received from the weak passer detection means.

In the above configuration, the third gate device (corresponding to claim 3) is preferably configured such that the pressure distribution data management means searches for an identification target area from all detectable target areas of the pressure distribution sensor. And an accumulated pressure distribution data updating means for obtaining accumulated pressure distribution data for the identification target area searched by the identification target area searching means.
According to said structure, the identification target area | region corresponding to a passer's footprint (shoe print), the contact trace of a cane tip part, the movement trace of a wheel, etc. is set, the pressure distribution data is calculated | required, Furthermore, the said pressure distribution data Is obtained by updating with a cumulative calculation at a required fixed time as the movement time elapses, so that pressure distribution data can be obtained with high accuracy, and the effectiveness of the subsequent determination can be enhanced.

In a fourth gate device (corresponding to claim 4), in the above-mentioned configuration, preferably, the weak passer-by detection means calculates Gaussian mixture distribution to obtain characteristic information of Gaussian mixture distribution from accumulated pressure distribution data related to the identification target region. And a distribution shape discriminating means for judging the distribution shape of the Gaussian mixture distribution using the characteristic information of the Gaussian mixture distribution provided from the Gaussian mixture distribution calculating means, and judging from this distribution shape whether or not the person is weak. It is characterized by having.
According to the above configuration, the footprint (bimodal Gaussian mixture distribution with respect to pressure distribution data) of a normal user in which only the footprint (shoeprint) remains, and the movement trace (pressure distribution) of a weak person using crutches, etc. The difference with respect to data (including unimodal Gaussian distribution) can be remarkably extracted and discriminated, and it can be quickly detected with high accuracy whether or not the person is a vulnerable person.

In a fifth gate device (corresponding to claim 5), in the above-mentioned configuration, preferably, the distribution shape discriminating means is configured such that when the Gaussian mixture distribution has a unimodal distribution shape, the passerby is a weak passerby. It is characterized by discriminating.

  According to the present invention, when a normal user and a passerby pass through a passage of a gate device such as an automatic ticket gate or a security gate, the pressure distribution for detecting the movement trace of the passer as pressure distribution data. A sensor, pressure distribution data management means for generating cumulative pressure distribution data relating to the identification target area based on the pressure distribution data, and whether or not the passer is a weak person based on the cumulative pressure distribution data relating to the identification target area. And a weak passage detection unit that gives a weak passage detection signal to the entry / exit control means when it is weak, and the entry / exit control means that receives the weak passage detection signal from the weak passage detection means Since the reading operation is stopped for a certain period, the distance between the vulnerable person and the normal user can be increased. Avoid contact and fall accident of a person, it is possible to secure a safe passage for the passage weak

It is a schematic plan view showing various patterns (A) to (E) such as footprints (shoe prints) according to passers-by who enter the ticket gate passage of the automatic ticket gate. It is a figure which shows the shape (A) of a pressure distribution in the case of two types of human footprints (shoe print), (B). It is a figure which shows the shape (A) and (B) of the pressure distribution in the case of traces other than a human footprint (shoe print) (wheel, cane tip). It is a figure for demonstrating the example of discrimination | determination of bimodality of pressure distribution by Gaussian mixture distribution, and monomodality. It is four state diagrams which consist of a set of a footer's footprint (shoe mark) mark and the pressure distribution corresponding to it. It is a figure explaining the example of transition of the pressure distribution at the time of normal walking. It is a block diagram which shows the structure of the traffic control system of an automatic ticket gate concerning this embodiment, and a weak traffic person discrimination | determination system. It is a flowchart which shows the flow of the process of the traffic control of an automatic ticket gate concerning this embodiment, and a weak person identification. It is a flowchart which shows the flow of a process of weak person detection.

  DESCRIPTION OF EMBODIMENTS Preferred embodiments (examples) of the present invention will be described below with reference to the accompanying drawings.

  An embodiment of an automatic ticket gate will be described as a representative example of a gate device according to the present invention. The automatic ticket gate is a machine that is installed at a ticket gate in a station premises and controls the passengers who get on and off the train on the railroad for the ticket gate. A traffic control system is provided in the automatic ticket gate. The automatic ticket gate according to the present embodiment further includes a system for determining whether a person (passenger) who wants to pass through the ticket gate is a normal person, that is, a normal user or a weak person. If the decision system determines that the passer is a weak passer, the pass control system changes the pass control compared to the pass control for a normal user, and the normal use that follows the passer. Increase the distance between the person and the person. For example, when it is determined that the person is weak, the IC card reader (IC card touch unit) or the magnetic ticket reader is not allowed to perform an accepting operation until a predetermined time has elapsed. This widens the distance between weak traffic users and normal users, thereby preventing collisions.

  Here, “passengers with weak traffic” refers to persons who have slower passage speeds and take longer to pass through the ticket gates than normal users, such as wheelchair users, crutch users, and bar staff users. It means that.

  A system for determining whether a passerby of a ticket gate is a weak person or a normal user performs discrimination (identification) of a passer by a position before the entrance of the ticket gate of an automatic ticket gate. The concept (principle) of the discrimination will be described with reference to FIGS.

  First, FIG. 1 will be described. FIG. 1 shows several patterns (A) to (E) regarding the footprints of various passers-by in the area just before the entrance 11A of the ticket gate passage 11 of the automatic ticket gate 10. The automatic ticket gate 10 is shown as a schematic plan view of two ticket gate bodies 10A and 10B installed in parallel, and a ticket gate 11 is formed between the two ticket gate bodies 10A and 10B in parallel. Yes. In FIG. 1, the lower end of the ticket gate passage 11 is set to the entrance 11A. The four patterns (B) to (E) on the left are shown as contrasts with the basic pattern (A).

  In FIG. 1, a pattern (A) is a pattern based on normal walking of a normal user, and is formed from equally spaced footprints (rear shoes) 21 by left and right shoes. The pattern (B) is a pattern based on the walking of a normal user having a carry back, and two linear trajectories 22 are generated by the wheels of the carry back together with the footprint 21 of the normal walking. Pattern (C) is a crutch user's pattern, and is formed of a footprint 21 on one foot and a trace 23 on the tip of left and right crutches. The pattern (D) is a pattern for a stick cane user, and is formed of left and right foot footprints 21 and a trace 24 at the tip of one stick cane. Pattern (D) is a wheelchair user's pattern, and is formed of a front wheel trajectory 25 and a rear wheel trajectory 26 having different left and right wheel intervals. As is clear from the above patterns (A) to (E), a passer-by is a normal walk (walking by a normal user), a normal user with a carry back, a crutch user, a stick cane user, Each wheelchair user has a different pattern. Therefore, if a mat sensor sensitive to pressure is arranged in the area 27 indicated by the solid line on the floor surface on the inlet 11A side of the ticket gate passage 11 of the automatic ticket gate 10, these patterns (A) to (E) are displayed on the mat sensor. Thus, it is possible to determine what type of user is the passer-by who intends to enter the ticket gate 11 before entering. In the following description, the specific area 27 is referred to as “mat sensor 27”.

  The mat sensor 27 having a rectangular planar shape capable of detecting the pressure distribution is configured by using a plurality of pressure sensors arranged in a lattice pattern inside the mat member. This is a known pressure distribution detection means. However, when used in the automatic ticket gate 10 according to the present embodiment, it is necessary to accurately detect vulnerable persons based on pressure distribution data on the upper surface taken out from the mat sensor 27 with high accuracy. Regarding the processing of the pressure distribution data supplied from the mat sensor 27 by the processing unit, the following processing is actually performed to determine whether the footprint is a shoe, a movement locus by a wheel, or a trace by a tip of a cane. Is done.

  In the pressure distribution data obtained from the mat sensor 27, when pressure data generated from something other than a human footprint (after the shoe) is detected, it is considered that the person is weak in principle. Even in the case of a normal user having a carry-back, it is actually expected that it will take longer to pass than a normal user having nothing. Therefore, it is desirable to use the same handling as those with weak traffic.

  The shape of the pressure distribution in the case of a human footprint (shoe print) is as shown in FIGS. 2A shows a footprint pressure distribution 32 by a female high heel 31, and FIG. 2B shows a footprint pressure distribution 34 by a normal shoe 33 or the like. In the illustrated pressure distributions 32 and 34, it means that the pressure is higher toward the lower side. In either case, when walking, the first part touches with the heel part, then the front part also makes contact, and finally moves so as to kick with the toe part. Accordingly, when the characteristics are grasped greatly in consideration of the passage of time, the pressure distribution is formed so that each of the pressure distributions 32 and 34 has two peak portions (or valley portions) 32a and 34a. .

  On the other hand, the shape of the pressure distribution in the case of traces other than human footprints (shoe prints) is as shown in FIGS. 3A shows the pressure distribution 36 due to the movement of the wheel 35 (arrow direction D1), and FIG. 3B shows the pressure distribution 38 due to the cane tip 37. FIG. In the pressure distribution 36, the pressure distribution is formed so as to have one gentle peak in the range of the rotational movement of the wheel 35, and in the pressure distribution 38, only a narrow portion where the cane tip 37 hits has one sharp peak. A pressure distribution is formed. When the pressure distributions 36 and 38 having such a single peak shape are detected, it can be considered that the user is a weak person.

  As described above, it is necessary to accurately identify and discriminate the shape of the pressure distribution based on the pressure distribution data detected by the mat sensor 27 in order to discriminate the traffic weak. In the case of this example, in the case of a normal user's footprint (shoe print), it is a pressure distribution having a multi-modal (two peaks) distribution shape. In addition to the distribution, a pressure distribution having a unimodal distribution shape is included. Accordingly, the signal (data) relating to the pressure distribution output from the mat sensor 27 is identified by determining whether or not the pressure distribution has a unimodal shape, thereby determining whether or not the person is weak. can do.

In the present embodiment, calculation processing for determining whether or not the pressure distribution shape has a single peak shape is performed in the following procedure.
A Gaussian mixture distribution model is used for the above discrimination. The Gaussian mixture distribution model is a distribution composed of a plurality of Gaussian distributions. When the Gaussian distribution is expressed by the following (Equation 1), the Gaussian mixture distribution is expressed by the following (Equation 2).

  In the above (Equation 2), K is the number of mixtures, πk is a mixture count, μk is an average, and Sk is a covariance matrix. In addition, since the Gaussian distribution in this case is two-dimensional, m = 2. When the number of mixing is 1, it becomes a single peak, and when it is 2, it becomes two peaks. Therefore, it is possible to examine whether or not it is a single peak by selecting the one having a higher degree of matching when the number of mixtures is 1 and when the number of mixtures is 2. The “degree of conformity” can be obtained from the degree of accuracy represented by the following equation (Equation 3).

  In the above (Equation 3), N represents the number of points (data number) sampled from the pressure distribution.

  FIG. 4 shows an example of determination of bimodality and unimodality of pressure distribution by Gaussian mixture distribution. 4A shows the shape of the pressure distribution obtained as the detection signal of the mat sensor 27. FIG. 4B shows a Gaussian mixture distribution obtained with a mixing number of 1 (K = 1), and FIG. 4C shows a mixing number of 2 (K = This is the Gaussian mixture distribution obtained in 2). In this case, the degree of conformity when the number of mixtures is 2 is high, and the pressure distribution in FIG. 4A is determined to be a bimodal shape, and is recognized as a pressure distribution due to human footprints. Similarly, when the pressure distribution obtained as the detection signal of the mat sensor 27 has a high degree of conformity when the number of mixtures is 1, the pressure distribution is determined as a single peak shape, and other than human footprints It is recognized that the pressure distribution is due to.

  Whether or not the pressure distribution obtained from the mat sensor 27 has a single peak shape is determined by determination based on the processing calculation using the Gaussian mixture distribution model as described above. Thus, when it is determined that the shape of the single peak is based on the pressure distribution obtained from the mat sensor 27 and the person who is going through the ticket gate passage 11 of the automatic ticket gate 10 is determined to be a weak person, the next pass of the weak person For the person, the operations of the IC card reader (IC card touch unit) and the magnetic ticket reader are temporarily stopped so that the passage interval is widened, and the ticket gate processing is delayed. By doing so, the next passer cannot pass near the weak passerby, so the possibility of a contact accident that causes a difference in pass speed can be greatly reduced.

  The pressure distribution signal output from the mat sensor 27 capable of detecting the pressure distribution generated on the predetermined plane area is converted into data, and the data relating to the pressure distribution is accumulated in the memory of the processing device. Cumulative pressure distribution data in a region where the pressure is zero for a certain period of time is to be determined. FIG. 5 shows an example of the accumulated pressure distribution.

  FIG. 5 shows four state diagrams (A) to (D) each including a set of a passer's footprint (actually a shoe print) mark 41 and a pressure distribution 42 detected by the mat sensor 27 corresponding thereto. ing. In each of the state diagrams (A) to (D) of FIG. 5, the vertical axis Y with respect to the footprint mark 41 means the direction in which the passerby moves upward, and the horizontal axis X with respect to the pressure distribution 42 is the pressure. It means size. Each position of the footprint mark 41 in the direction of the vertical axis Y corresponds to the pressure value in the direction of the vertical axis Y of the pressure distribution. In addition, the state diagrams (A), (B), and (C) sequentially change as time (t) elapses.

  In the state diagram (A) of FIG. 5, the shoe heel first comes into contact with the surface of the mat sensor 27 (the footprint mark 41 of the heel) and pressure is applied, and a pressure distribution 42 corresponding to this is generated. In the state diagram (B), the entire back surface of the shoe is in contact with the surface of the mat sensor 27 (the footprint mark 41 on the entire back surface), and a pressure distribution 42 corresponding to the pressure is generated. In the state diagram (C), the passerby has further advanced, and only the front half (toe portion) of the back of the shoe is in contact with the surface of the mat sensor 27 (the footprint mark 41 in the front half) and pressure is applied. A pressure distribution 42 corresponding to this is generated. As a result of the above state diagrams (A) to (C), the cumulative pressure distribution generated in the footprint mark 41 in each state is shown in the state diagram (D). In the state diagram (D), the footprint mark disappears, the passerby has already moved forward, and the cumulative pressure distribution 42 is the pressure distribution at that time shown in the state diagrams (A) to (C). 42 is added. In this way, the pressure distribution 41 generated in the series of footprint marks 41 at the same location is acquired as cumulative pressure distribution data (pressure distribution 42 in the state D) at the location, and can be identified by this. .

Next, an example of transition of pressure distribution during normal walking will be described with reference to FIG. More specifically, FIG. 6 shows the transition of the footprint mark 41 at the time transition (t _{1 to} t ₇ ) of the seven stages in the mat sensor 27, that is, the transition of the pressure distribution. Time t ₁ is a state in which the left foot arrived, the time t ₂ is a state in which the heel arrived of the right foot, the time t ₃ is a state in which the toes of the right foot had to leave the heel of the left foot and at the same time ground. The time t ₄ is the left foot is in the completely separated state, the pressure of the portion left foot is in contact with the ground is zero. After a certain period of time when the pressure becomes zero, the pressure distribution data in this portion is identified (determined) using the accumulated pressure distribution data described above. The accumulated pressure distribution data in the identification target area is cleared after identification. Identify after a certain time has elapsed, for example, in order to avoid identification of a state such as at time t _3. Considering pressure measurement errors, if the pressure is very small, it may be considered that the pressure is zero. Further, the transition of the pressure distribution from time t _{3 to} t ₇ is the transition of the footprint mark on the right foot. The same applies to the right foot.

  Next, with reference to FIGS. 7-9, the structure regarding the traffic control system of the automatic ticket gate 10 which concerns on this embodiment, and its operation | movement are demonstrated. This traffic control system incorporates a system for determining whether or not the passerby of the ticket gate is a passerby weak person (passage weakness discrimination system). In this determination system, the principle of determination as to whether or not the passerby described with reference to FIGS. 1 to 6 is a weak passerby is applied.

  In FIG. 7, the automatic ticket gate 10 includes a ticket gate controller 51 for performing a conventional ticket gate control known in the art. The ticket gate control unit 51 normally receives ticket gate information on passers-by read by the IC card reader 52 or the magnetic ticket reader 53, determines whether it is valid / invalid, controls the opening / closing of the passage prevention door, and charges Execute processing etc. At this time, the ticket gate control unit 51 controls the receiving operation of the IC card reader 52 and the magnetic ticket reader 53 on the assumption that the passerby is a normal user. The reception operation of the IC card reader 52 is wireless transmission / reception with the IC card carried by the passer-by, and the reception operation of the magnetic ticket reader 53 sets the magnetic ticket insertion slot lid 53a to the open state and the passer-by This is an operation that enables insertion of a magnetic ticket possessed by the machine into the slot.

  In the automatic ticket gate 10 according to the present embodiment, as shown in FIG. 7, a pressure distribution sensor 61, a pressure distribution data management unit 70, and a traffic weakness detection unit 80 are further provided for the above-described ticket gate control unit 51. . The pressure distribution sensor 61, the pressure distribution data management unit 70, and the traffic weak detection unit 80 constitute a traffic weak discrimination system.

  The pressure distribution sensor 61 corresponds to the mat sensor 27 described above. The pressure distribution sensor 61 is constituted by a plurality of pressure sensors arranged in a grid pattern in a planar area of the mat surface. The pressure distribution data output from the pressure distribution sensor 61 is input to the pressure distribution data management unit 70.

  The pressure distribution data management unit 70 includes a cumulative pressure distribution data update unit 71, a memory 72, and an identification target region search unit 73. The pressure distribution data provided from the pressure distribution sensor 61 is input to the cumulative pressure distribution data update unit 73. The accumulated pressure distribution data is updated by the accumulated pressure distribution data updating unit 71 and then stored in the memory 72 continuously. The accumulated pressure distribution data update unit 71 reads the accumulated pressure distribution data stored in the memory 72 every time the pressure distribution data is input, performs further accumulation calculation processing, and then stores the accumulated pressure distribution data in the memory 72 again. The identification target area search unit 73 designates an identification target area on the pressure receiving surface of the mat-shaped pressure distribution sensor 61, and takes in the accumulated pressure distribution data related to the identification target area from the cumulative pressure distribution data update unit 71. The identification target area search unit 73 sends the accumulated pressure distribution data of the identification target area to the weak vulnerable person detection unit 80 after a predetermined condition.

  The weak traffic detection unit 80 includes a Gaussian mixture distribution calculation unit 81 and a distribution shape determination unit 82. The Gaussian mixture distribution calculation unit 81 takes in the accumulated pressure distribution data of the identification target region and obtains the aforementioned Gaussian mixture distribution parameter. The Gaussian mixture distribution calculation unit 81 provides the obtained Gaussian mixture distribution parameter and the accumulated pressure distribution data of the identification target region to the distribution shape determination unit 82. The distribution shape determination unit 82 determines a Gaussian mixture distribution corresponding to the Gaussian mixture distribution parameter based on the degree of matching, and determines whether or not the accumulated pressure distribution data in the identification target region is a single peak. When it is determined that the accumulated pressure distribution data in the identification target area is unimodal, it is determined that the passer is a passerby person, and a weak passer detection signal is output from the distribution shape determiner 82 to the ticket gate controller 51. The

  When the ticket gate controller 51 receives the vulnerable person detection signal from the distribution shape discriminating part 82 of the vulnerable person detection part 80, the ticket gate controller 51 accepts it to the IC card reader 52 and the magnetic ticket slot cap 53a of the magnetic ticket reader 53. Provide a stop signal. When receiving the reception stop signal, the IC card reader 52 stops the reception operation. When receiving the acceptance stop signal, the magnetic ticket reader 53 operates the magnetic ticket slot cover 53a to close the magnetic ticket slot, thereby stopping the acceptance operation. After that (for example, after about 2 to 3 seconds), the ticket gate control unit 51 provides an acceptance start signal to the IC card reader 52 and the magnetic ticket slot cap 53a of the magnetic ticket reader 53 to stop the acceptance. Release and start accepting operation. Thus, when the next passer following the weak passer is a normal user, the ticket gate control is performed to widen the distance between the passerby and the normal user by slightly delaying the reception operation.

  Next, a flow of a series of processes related to the above operation will be described with reference to the flowcharts of FIGS.

  In FIG. 8, the determination step S11 is executed in the first step. The determination step S11 determines whether or not the pressure distribution sensor 61 has detected an object, in other words, whether or not pressure due to a load of a passerby or the like trying to pass through the ticket gate passage 11 of the automatic ticket gate 10 is detected. To do. Specifically, in FIG. 7, this is determined based on whether or not the pressure distribution data is provided from the pressure distribution sensor 61 in the accumulated pressure distribution data update unit 71 of the pressure distribution data management unit 70. . When the determination step S11 is NO, the subsequent steps are not executed, the end state is reached, and the determination step S11 is periodically repeated at a predetermined time interval. If YES in the determination step S11, the process proceeds to the next processing step S12.

  In processing step S12, the accumulated pressure distribution data is updated. This processing content is executed by the cumulative pressure distribution data update unit 71 that has received the pressure distribution data from the pressure distribution sensor 61. Thereafter, the process proceeds to the next determination step S13.

  The determination step S13 is a step for determining whether or not a normal ticket gate process has been performed, and it is determined whether or not the IC card reading process or the magnetic ticket reading process has been completed. If NO in the determination step S13, the process returns to the first determination step S11 again, and the determination step S11 is executed. This is to confirm whether or not an object (passerby) is actually detected. If NO in the determination step S11 again, the process ends as a detection error or the like. If YES in determination step S13, processing routine S100 for detecting a passerby person is executed. The contents of the processing routine S100 for detecting a weak person are the same as described in the detection principle and the pressure distribution data management unit 70 and the weak person detection unit 80, and this is shown in FIG. It becomes such a flowchart.

  In FIG. 9, an identification target region is first searched (processing step S101), and a Gaussian mixture distribution is calculated based on the pressure distribution data relating to the searched identification target region (processing step S102). Next, it is detected that the person is a traffic weak person based on the Gaussian mixture distribution (processing step S103), and thereafter the accumulated pressure distribution data relating to the identification target region is further cleared (processing step S104). In this way, the processing routine S100 for detecting a vulnerable person is executed, and a process for detecting a vulnerable person is performed.

  In the flowchart of FIG. 8 again, in the next determination step S14, it is determined whether or not a weak person has been detected as a result of the weak person detection (routine S100). If no passerby is detected (in the case of NO), the process is terminated, and the ticket gate control unit 51 executes normal ticket gate control. When a passerby is detected (in the case of YES), the ticket gate controller 51 gives an acceptance stop signal to the IC card reader 52 and the magnetic ticket slot cover 53a of the magnetic ticket reader 53 to read the IC card. The operation or reading operation of the magnetic ticket is stopped (processing step S15). Thereafter, the reading operation is stopped for a predetermined time (processing step S16). After the arbitrary fixed time has elapsed, the ticket gate controller 51 gives an acceptance start signal to the IC card reader 52 and the magnetic ticket slot 53a of the magnetic ticket reader 53 to read the IC card or read the magnetic ticket. The operation is started and the accepting operation is resumed (processing step S17).

  When a passerby enters the ticket gate passage 11 of the automatic ticket checker 10 by the above processing steps S15 to S17 related to the ticket checker control unit 51, the acceptance operation for the ticket gate is subsequently delayed, and the passerby is followed. Thus, the entry operation of the user who tries to enter the ticket gate passage 11 is delayed, and the interval between the weak person and the successor is widened. As a result, it is possible to eliminate the accident of contact and collision between the weak traffic person and the subsequent normal user.

  In the above embodiment, an example of an automatic ticket gate has been described. However, the present invention is not limited to an automatic ticket gate, and can be generally applied to a gate device such as a security gate. In the case of a general gate device, the above-mentioned ticket gate passage serves as a passage for entry and exit, and the IC card reader 52 and magnetic ticket reader 53 have an entrance / exit medium possessed by a passerby who wants to pass through the passage. It becomes an entrance / exit information reading device that transmits / receives traffic information to / from (entrance IC cards, magnetic tickets, etc.). The ticket gate control unit 51 serves as an entrance / exit control means for controlling the reading operation of the entrance / exit reader.

  The configurations, shapes, sizes, and arrangement relationships described in the above embodiments are merely schematically shown to the extent that the present invention can be understood and implemented, and numerical values are merely examples. Therefore, the present invention is not limited to the described embodiments, and can be variously modified without departing from the scope of the technical idea shown in the claims.

  The gate device according to the present invention can quickly detect a vulnerable person with high accuracy when a normal user (passenger) or a vulnerable person passes through an automatic ticket gate installed at a ticket gate of a station. By doing so, when a normal user tries to pass following a weak person, it is used to prevent accidents such as contact by delaying the ticket gate control process a little.

10 Automatic ticket gates 10A, 10B Ticket gate body 11 Ticket gate 11A Entrance 21 Footprints (shoe marks)
22 Trajectory of wheel 23 Trace of tip of crutch 24 Trace of tip of crutch 25, 26 Trajectory of front and rear wheels of wheelchair 27 Area (mat sensor)
31 High heel 33 Shoes 32, 34 Pressure distribution 35 Wheels 37 Cane tip 36, 38 Pressure distribution 51 Ticket gate control unit 52 IC card reader 53 Magnetic ticket reader 61 Pressure distribution sensor 70 Pressure distribution data management unit 80 Traffic weak detection unit

Claims (5)

A pressure distribution sensor that detects a person's movement trace as pressure distribution data;
Pressure distribution data management means for generating cumulative pressure distribution data related to the identification target region based on the pressure distribution data output from the pressure distribution sensor;
Control means for determining the characteristics of the passer based on the accumulated pressure distribution data, and performing pass control according to the characteristics ;
A gate device comprising: An entrance / exit information reader for transmitting / receiving traffic information to / from an entrance / exit medium possessed by a passerby passing through the passageway;
Entrance / exit control means for controlling the reading operation of the entrance / exit reading device ;
It is determined whether the the passer is passing weak on the basis of the cumulative pressure distribution data according to the recognition target area supplied from the previous SL pressure distribution data management unit, complete competed control means before when the a traffic vulnerable A traffic weak detection means for giving a traffic weak detection signal to
2. The gate device according to claim 1, wherein the entrance / exit control means stops the reading operation of the entrance / exit reading device for a certain period when the vulnerable person detection signal is received from the vulnerable person detection means. The pressure distribution data management means includes: an identification target area search means for searching the identification target area from all detectable target areas of the pressure distribution sensor; and the accumulated pressure for the identification target area searched by the identification target area search means. the gate apparatus of claim 1, wherein it has a cumulative pressure distribution data updating means for obtaining the distribution data.   The weak traffic detection unit is a Gaussian mixture distribution calculating unit that obtains characteristic information of a Gaussian mixture distribution from the accumulated pressure distribution data related to the identification target region, and the Gaussian mixture distribution provided by the Gaussian mixture distribution calculating unit. 3. The gate device according to claim 2, further comprising distribution shape determining means for determining a distribution shape of the Gaussian mixture distribution using feature information and determining whether the Gaussian mixture distribution is weak or not.   5. The gate device according to claim 4, wherein the distribution shape determining means determines that the passer is a passer when the distribution shape of the Gaussian mixture distribution has a unimodality.

Images