JP2011128107A - Device for controlling moving body - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a device for controlling a moving body, capable of exactly identifying a moving body without a tag. <P>SOLUTION: A traffic line producing section 2 of a moving body with ID produces the spatiotemporal traffic line of a tag based on the detection result of position detecting sections 1a, 1b, and 1c of the moving body with ID as a traffic line of the moving body with ID. A traffic line producing section 4 of a moving body produces the spatiotemporal traffic line of the moving body based on the detection result of position detecting sections 3a, 3b, and 3c of the moving body as a traffic line of the moving body. A traffic line comparing section 5 compares the traffic line of the moving body with ID and the traffic line of the moving body to make moving bodies to match which have closer traffic lines. An identifying section 6 identifies the moving body having no corresponding traffic line as the moving body without a tag in the traffic line comparing section 5. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention detects, manages, and distinguishes the position of a mobile object such as a person having an ID-identifiable device (hereinafter referred to as a tag) and the position of a mobile object such as a person who does not have an ID-identifiable device. The present invention relates to a body management device.

  Conventionally, in a monitoring system that detects and distinguishes between a person with a tag and a person without a tag, for example, in a local area, a person without a tag is found by comparing the number of tags with the number of persons. (For example, refer patent document 1). In addition, there is a management device that detects anomalies such as tag switching by comparing behavior histories of different types of individuals and finding inconsistencies in those behaviors (see, for example, Patent Document 2). .

JP 2008-16042 A JP 2008-293137 A

  However, in the monitoring system as described in Patent Document 1, since the number of people and the number of tags are compared in the local area, if a detection error occurs in the local area, the person who does not have the tag There was a problem that it was difficult to discover. Further, even if a tag non-holder is detected, there is a problem that the tag non-holder may be lost if there is no means for breaking the exit path. In addition, since the management device described in Patent Document 2 compares only behavior histories obtained using tags, there is a problem that the behavior of a person who does not have a tag cannot be detected.

  The present invention has been made to solve the above-described problems, and an object of the present invention is to obtain a moving body management apparatus that can accurately determine a moving body having no tag.

  A mobile body management apparatus according to the present invention is a mobile body management apparatus for identifying a mobile body having a tag and a mobile body having no tag, and a mobile body position detection unit with ID for detecting the position of the tag, A moving body flow line creation unit with ID for creating a moving line of the space time of the tag as a moving body flow line with ID from the detection result of the moving body position detection unit with ID, a moving body position detection unit for detecting the moving body, The moving body flow line creation unit that creates a moving line in the spatio-temporal space of the moving body from the detection result of the moving body position detection unit is compared with the moving body flow line with ID and the moving body flow line. A flow line comparison unit for associating objects that are close to each other and a determination unit that determines, in the flow line comparison unit, a moving body of a flow line that is not associated as a moving body that does not have a tag.

  The moving body management apparatus according to the present invention determines a moving body that does not have a tag by comparing the flow line of the moving body having a tag with the flow line of the moving body that is detected regardless of the presence or absence of the tag. Therefore, it is possible to accurately determine a mobile object that does not have a tag.

BRIEF DESCRIPTION OF THE DRAWINGS It is a block diagram which shows the mobile body management apparatus by Embodiment 1 of this invention. It is a flowchart which shows operation | movement of the mobile body flow line creation part with ID of the mobile body management apparatus by Embodiment 1 of this invention. It is explanatory drawing of the data integration of step ST202 of the moving body management apparatus by Embodiment 1 of this invention. It is explanatory drawing of flow line creation of step ST203 of the moving body management apparatus by Embodiment 1 of this invention. It is a flowchart which shows operation | movement of the mobile body flow line preparation part of the mobile body management apparatus by Embodiment 1 of this invention. It is explanatory drawing which shows the example of a detection in the mobile body position detection part of the mobile body management apparatus by Embodiment 1 of this invention. It is explanatory drawing which shows the operation | movement produced in the mobile body flow line preparation part of the mobile body management apparatus by Embodiment 1 of this invention. It is explanatory drawing which compares the human flow line with ID and a human flow line for every time interval with the mobile body management apparatus by Embodiment 1 of this invention. It is explanatory drawing which compares the human flow line with ID and a human flow line for every area section by the mobile body management apparatus by Embodiment 1 of this invention. It is explanatory drawing in the case of calculating | requiring a human presence rate from the human flow line with ID of the mobile body management apparatus by Embodiment 2 of this invention. It is explanatory drawing of the person presence rate calculated | required in the mobile body flow line preparation part of the mobile body management apparatus by Embodiment 2 of this invention. It is explanatory drawing which shows the example of arrangement | positioning of the detection part of the mobile body management apparatus by Embodiment 5 of this invention. It is a block diagram in the case of performing the parallel process of the mobile body management apparatus by Embodiment 7 of this invention. It is a block diagram in the case of using the comparison log | history of the mobile body management apparatus by Embodiment 7 of this invention. It is a block diagram in the case of changing the detection parameter of the moving body management apparatus by Embodiment 8 of this invention.

Embodiment 1 FIG.
FIG. 1 is a block diagram showing a mobile management apparatus according to Embodiment 1 of the present invention.
The mobile body management apparatus shown in FIG. 1 includes ID-attached mobile body position detection units 1a, 1b, and 1c, ID-attached mobile body flow line creation unit 2, mobile body position detection units 3a, 3b, and 3c, and mobile body flow line creation unit. 4, a flow line comparison unit 5 and a determination unit 6 are provided.

  The ID-attached moving body position detection units 1a, 1b, and 1c are functional units that detect the positions of tags held by a person who is a moving body. The ID-attached mobile body position detection units 1a, 1b, and 1c include, for example, a receiver that reads an RF-ID tag, and reads ID information of a tag that exists near the receiver. Then, by linking the read tag ID with the receiver ID, the reading time, and the receiver installation position information, it is detected who is where and when. In addition to the RF-ID, the ID of the tag may be read using human body communication, visible light communication, infrared communication, GPS, wireless LAN, UWB, or the like to specify the position of the person. Further, the ID-attached moving body flow line generating unit 2 uses the ID-attached moving body position detecting units 1a, 1b, and 1c as the ID and the time-space flow line of the tag as the ID-added person moving line (ID-attached moving body flow line). It creates a flow line and action history such as stay time and movement history for each ID.

  The moving body position detection units 3a, 3b, and 3c are functional units that detect a person as a moving body regardless of whether or not the tag is held. In other words, these mobile body position detection units 3a, 3b, and 3c are devices that detect the presence and movement of a person, such as pyroelectric sensors, temperature sensors, ultrasonic sensors, pressure sensors, distance sensors, and cameras. Although it cannot be specified, it is a device that detects a person regardless of the tag holding status. For example, in the case of a human sensor using a pyroelectric sensor, since a change in infrared rays emitted by a person is detected, it is possible to detect a person's movement such as crossing the sensor detection range. In addition, by measuring the distance to the floor from time to time with an ultrasonic sensor attached to the ceiling and storing the distance at the time when no person is present as the reference distance, there is a person at the time when the difference from the reference distance is large You may detect that you are doing. Further, a mat-shaped pressure sensor may be laid on the floor, and a person may be present in a portion where pressure is applied. In addition, the moving body flow line creation unit 4 is a functional unit that creates a human existence probability, a human flow rate, a number distribution, and the like using detection results of the moving body position detection units 3a, 3b, and 3c. That is, the moving body flow line creation unit 4 calculates the moving direction, the moving speed, the approximate number of people, and the like by using the position-related information / detection history with the surrounding sensors. Output as flow line information).

  The flow line comparison unit 5 compares the ID-added human flow line information created by the ID-attached mobile flow line creation unit 2 with the human flow line information created by the mobile body flow line creation unit 4, and the flow lines are close to each other. It is a functional part that associates things. For example, when comparing human flow line information with an ID and human flow line information with an ID, there is a person in the human flow line information, but in the human flow line information with an ID, a person with any ID is in that area. If there is no result, it can be seen that there were people who did not have tags in the area. The determination unit 6 is a functional unit that obtains the length of a human flow line that has not been matched in the flow line comparison unit 5 and determines that the length is equal to or greater than the time threshold or the distance threshold as a tag-free person. .

  Hereinafter, when a receiver that reads an RFID tag is used as the mobile position detectors 1a, 1b, and 1c with ID, and a human sensor using a pyroelectric sensor is used as the mobile body position detectors 3a, 3b, and 3c, the tag An example of detecting a holder will be described.

  The RFID receiver reads the ID of a tag within the detection range of the receiver, and outputs the receiver number, the ID number of the read tag, and the reading time. The human sensor detects the amount of change in infrared light within the detection range, determines that there is a moving person if the amount of change exceeds a certain value, and outputs the human sensor number, motion detection result, and detection time . When the tag detection timing of the receiver and the human detection timing of the human sensor are equal and the detection range of the receiver and the human sensor is the same, only the output of one receiver and the output of one human sensor are compared. Thus, it is possible to detect a person who does not have a tag. For example, when the human sensor detects a person at a time when the receiver does not detect the tag, the receiver determines that the person who does not hold the tag is within the detection range.

  Thus, by comparing the output results of one receiver and one human sensor, it is possible to determine the presence / absence of an approximate person and the tag holding status, but the RFID receiver and human sensor detect detection information. There are patterns that cannot be detected, such as when the tag holder is stationary, or when the tag holder and the holder do not enter the detection range at the same time. Also, it is difficult and unrealistic to adjust the detection range of the RFID receiver and the human sensor to be equal. Further, since the detection timing is different between the RFID receiver and the human sensor, the detection time may be shifted. For example, when the human sensor detects a movement earlier than the RFID tag receives the ID, it may be erroneously determined that a person who has no tag is detected.

  In particular, when detecting and managing the positions of tag holders and non-holders throughout a building such as a building, a very large number of devices must be disposed, and ID-attached mobile body position detection units 1a, 1b, and 1c. It is difficult to adjust all the detection ranges and detection timings of the mobile body position detection units 3a, 3b, and 3c. Therefore, in the present invention, instead of directly comparing the detection results, the ID-added moving body flow line creating unit 2 creates the ID-added human flow line, and the moving body flow line creating unit 4 creates the human position flow line, By comparing these, non-tag holders are detected.

  Next, creation of an ID-added human flow line in the ID-attached mobile flow line creation unit 2 will be described with reference to FIG. In step ST201, tag detection results from the ID-attached mobile body position detection units 1a, 1b, and 1c are obtained. In step ST202, tag detection results with the same ID that are close in space and time are integrated. In step ST203, a flow line for each ID is created by connecting the tag detection data integrated in step ST202 in time series.

  Next, data integration in step ST202 will be described with reference to FIG. When a plurality of receivers detect a tag with the same ID at the same time, or when there is data with a similar detection time, the detection data is integrated. Here, it is assumed that the RFID tag receivers 1a and 1b are installed at the position coordinates xa and xb as the ID-attached mobile body position detection units 1a and 1b, respectively. In the figure, detection histories 301a and 301b are detection results of the tags of the receivers 1a and 1b, and t2 <t3 <t4. The start time of the integration time range for performing data integration is t1, the end time is t5, and the detection data with the same ID is integrated among the detection data included in the times t1 to t5. If there are multiple receivers in the integrated data and the position is below the threshold value, the detection result after integration is tag position = Σ receiver position coordinates / number of receivers, detection time = (start integration) Time + Integration end time) / 2.

  When the distance between the receiver positions xa and xb is equal to or less than the threshold value and t1 <t2 <t3 <t4 <t5, the detection results of the three tag IDs 001 at times t2 to t4 are integrated. Detection time) = (xa + xb) / 2, (t1 + t5) / 2. When the distance between the receiver positions xa and xb is equal to or less than the threshold value and t1 <t2 <t3 <t5 <t4, the detection results of the two tag IDs 001 at times t2 and t3 are integrated. Detection time) = (xa + xb) / 2, (t1 + t5) / 2, and then the tag 001 is detected at (xa, t4). When the distance between the receiver positions xa and xb is equal to or greater than the threshold and t1 <t2 <t3 <t4 <t5, the data of t2 and t4 are integrated, and (detection position, detection time) = xa, (t1 + t5) / 2 The threshold for the integrated distance is determined from the detection range of the receiver, and the integrated time range is determined from the reading cycle of the receiver. The integration distance threshold and the integration time range may be changed for each position. The tag position may be Σ weight * receiver position coordinate / number of receivers, and the weight may be obtained from the radio wave intensity at the time of reception.

  Next, the flow line creation in step ST203 will be described with reference to FIG. It is assumed that the detection history of ID001 after data integration in step ST202 is 401, and t1 <t2 <t3 <t4 <t5 <t6 <t7 <t8. When the pairs of detection positions and detection times of the tag ID 001 are arranged in time series, they become (xa, t3), (xa, t5), (xb, t8), and this data is plotted on the coordinates and connected in a time series with a straight line. Then, the flow line of ID001 becomes like the line of the graph 402. Here, the position coordinates are not one-dimensional but may be two-dimensional or three-dimensional. A flow line is similarly created for other IDs. Here, there is also a method of creating more accurate flow line information using information on immovable areas such as walls and obstacles as the spatial information. If the flow line collides with the immovable area when connecting the tag detection results with the flow line, the shortest path that does not collide with the immovable area is taken. The collision between the immovable area and the flow line and the derivation of the shortest path are calculated using geometry.

Next, a method for creating a human flow line in the moving body flow line creation unit 4 will be described with reference to FIG.
In step ST501, tag detection results are obtained from the moving body position detectors 3a, 3b, 3c. In step ST502, detection results that are close in space and time are integrated. In step ST503, the detection data integrated in step ST502 is tracked and associated. In step ST504, a flow line is created by connecting the detection data associated in step ST503.
The data integration method in step ST502 is the same as the data integration method in step ST202. However, since ID information cannot be obtained in step ST501, the target of data integration in step ST502 is not for each ID but for all detection data.
In step ST503, association is performed in order to create a flow line by connecting the detection data after the data integration in step ST502 in step ST504.

The association method in step ST503 will be described with reference to FIGS.
As shown in FIG. 6, as the moving body position detection units 3a, 3b, and 3c, human sensors 3a, 3b, and 3c are installed at position coordinates xa, xb, and xc, respectively, and each has a detection range and an influence range. Suppose you are. The detection histories 601a, 601b, and 601c are the motion detection results of the human sensors 3a, 3b, and 3c, and have been integrated in step ST502. Here, t1 <t2 <t3 <t4 <t5 <t6 <t7 <t8 <t9. If a human sensor is used as the moving body position detection unit 3a, 3b, 3c, a still person cannot be detected. Therefore, there is a person within the detection range of the human sensor that detects the movement and stays at that position even after detection. Suppose you continue. Also, if a person is present in the vicinity area when the motion sensor detects the movement, it is assumed that the person has moved to the position of the sensor that detected the movement. When there are two or more people in the vicinity area, the closer one is associated. If the distance between xa and xb and the distance between xb and xc are less than or equal to the neighborhood threshold, and the distance between xa and xc is greater than or equal to the neighborhood threshold, the association is performed based on the detection history 601a, 601b, and 601c. The result is a human flow line 701 as shown in FIG.

  When the positions and times of the human sensors that have detected the movement are arranged in chronological order, (xa, t3), (xb, t6), (xb, t7), and (xc, t8) are obtained. After the human sensor 3a detects the movement at time t3, a person stays at the position xa. When the human sensor 3b detects a movement at time t6, since the distance between xa and xb is equal to or smaller than the neighborhood region threshold, it is determined that the person has moved from the position xa to the position xb. Therefore, the stay at the position xa is canceled and a person stays at the position xb. Similarly, when the human sensor 3c detects a movement at time t8, the person staying at the position xb has moved to the position xc. Even if the human sensor 3a detects a movement at time t9, the distance between xa and xc is equal to or greater than the neighborhood region threshold, so that the position of the person staying at the position xc does not move to the position xa, and the position xa It is determined that a new person has started to stay. The threshold value of the neighborhood region is determined from the detection range of the human sensor, the installation interval, and the like. This threshold may be changed depending on the position. As an association method, tracking may be performed using a time series filter such as a Kalman filter or a particle filter.

  The flow line creation in step ST504 may be discrete like the human flow line 701 in FIG. 7, but the position movement between the sensors is smoothly interpolated using the average walking speed of the person. Also good. The interpolated flow line is shown as a human flow line 702. If it is determined that there is a movement from the position xa to xb at time t6, the movement start time t10 is calculated from the average walking speed of the person and the distance between xa and xb, and (xa, t10) and (xb, A flow line is interpolated by connecting t6) with a straight line. However, when the movement start time t10 exceeds the final detection time t3 at the position xa, the final detection time is set as the movement start time. In the movement from xb to xc resulting from the detection at (xc, t8), when the movement start time t11 at xb is calculated, t11 <t7, so the movement start time is t7, and (xb, t7) and (xc , T8). If the position movement between the sensors is smoothly interpolated, a more accurate flow line can be created.

Next, the flow line comparison of the flow line comparison unit 5 and the determination process of the determination unit 6 will be described.
The flow line comparison unit 5 compares the flow lines of the ID-added human flow line created by the ID-attached mobile flow line creation unit 2 with the human flow line created by the mobile body flow line creation unit 4. That is, the flow line comparison unit 5 obtains the human flow line corresponding to the flow line with ID for all the human flow lines with ID, and attaches the label ID to the corresponding human flow line. Further, the determination unit 6 obtains the length of the human flow line on which the corresponding label is not pasted, and determines that the length is equal to or greater than the time threshold or the distance threshold as a tag non-holder.

  Next, an association method in the flow line comparison unit 5 will be described. For the association, the distance between the movement vectors of the flow line is used, and when the distance between the movement vectors is equal to or less than the threshold value, it is assumed that the association is successful, and a label ID is attached to a part of the human flow line. Here, the movement vector is represented by a set of start point (start point position, start point position existing time) → end point (end point position, end point position existing time), and is obtained by cutting a flow line.

  A method for calculating the distance of the movement vector will be described with reference to FIG. The flow line 801 of the tag IDi is cut at time intervals t1 to t2, and a movement vector 802 between the times [t1, t2] of IDi is obtained. When the human flow line 803 is cut out at time intervals t1 to t2, a human movement vector 804 and a human movement vector 805 are obtained. Next, the distance between the IDi movement vector 802 and all human movement vectors is obtained, and the distances that are equal to or smaller than the threshold are associated with each other. Since calculating the distances to all the human movement vectors requires a calculation cost, the vicinity area may be searched and the distance from the unlabeled person movement vector 804 in the vicinity area may be calculated. The distance is calculated by, for example, the sum ΣD (xit, xt) of the position at each time (xit is the position at the time t of the IDi movement vector, xt is the position at the time t of the human movement vector, and D (a, b) is Euclidean distance between two points a and b) is used. The distance between vectors may be obtained by other methods such as using Pearson's correlation coefficient.

  A method for calculating the distance between different movement vectors will be described with reference to FIG. When the flow line 901 of IDi is cut out in the area section 905, a movement vector 902 of IDi is obtained. When the human flow line 903 is cut off in the time interval [t1-Δt, t2 + Δt] around the area interval 905 and the IDi movement vector 902, a human movement vector 904 is obtained. Next, the distance between the movement vector 902 of IDi and the person movement vector 904 is calculated. Since there is a difference in detection timing and installation position between the flow line with ID and the human flow line, the distance of the movement vector is calculated using the speed and the movement direction (angle). For example, it is assumed that distance = w1 * angle difference + w2 * speed difference (w1 and w2 are weights), and those that are equal to or less than the threshold are associated.

  In the examples of FIGS. 8 and 9, the local flow line comparison method has been described. However, since local comparison alone is likely to cause an error, the vector similarity is obtained by increasing the time or area interval. Also good. The similarity of the human flow line is obtained along the flow line with ID, and the label ID may be pasted when the similarity is high in a time or a length longer than a threshold in time or space.

  In the above explanation, an example of detecting tag non-holders for the purpose of monitoring was explained. However, it is not possible to grasp and distinguish the flow line and human distribution of both tag holders and tag non-holders. It can also be applied to services. For example, a service that allows a guard to have a tag and grasps the flow line, detects the position of a general person as a tag-free person, and dynamically sets the patrol route of the guard based on the positional relationship between the two Applicable to. Also, at commercial facilities, etc., the store clerk is given a tag and the position of the store clerk is managed based on the tag position. It can be applied to services such as detecting the movement of ordinary people and dispatching the nearest store clerk.

  Thus, in the first embodiment, even if the detection ranges and detection timings of the ID-attached mobile body position detection units 1a, 1b, and 1c and the mobile body position detection units 3a, 3b, and 3c are different, the ID-attached mobile body flow line is provided. By obtaining the respective flow lines by the creation unit 2 and the moving body flow line creation unit 4, the flow line comparison unit 5 can compare and detect a tag non-holder. In particular, when sensors are arranged to detect the entire building such as a building, a great number of devices must be arranged, and the ID-attached mobile body position detection units 1a, 1b, 1c and the mobile body position Since it is difficult to adjust all the detection ranges and detection timings of the detection units 3a, 3b, and 3c, this method is effective.

  In addition, since the flow line is created, it is possible to track the non-tag holder after detecting the non-tag holder and notify the current position of the non-tag holder. Furthermore, it is possible to examine the position history of the tag non-holder by tracing the flow line in the past direction and check the time when the tag non-carrier entered the video surveillance area. You can get a video. In addition, since it is possible to check the ID number of the tag that the non-tag holder has met or moved together, the ID number that asks the tag non-holder for the presence of a suspicious person, is likely to be accompanied You can specify.

  As described above, according to the mobile object management apparatus of the first embodiment, the mobile object management apparatus identifies a mobile object that has a tag and a mobile object that does not have a tag, and an ID that detects the position of the tag. ID-attached mobile object that creates a time-space flow line of a tag as an ID-attached mobile object flow line from the detection results of the ID-attached mobile object position detectors 1a, 1b, 1c and the ID-attached mobile object position detectors 1a, 1b, 1c From the detection results of the flow line creation unit 2, the mobile body position detection units 3a, 3b, and 3c that detect the mobile body, and the mobile body position detection units 3a, 3b, and 3c, In the moving body flow line creation unit 4 to create as a line, the moving body flow line with ID and the moving body flow line, and the flow line comparison unit 5 and the flow line comparison unit 5 that associate the movement line close to each other, A moving body with a flow line that is not associated is determined as a moving body without a tag. Since a determination section 6 that can be determined accurately mobile no tags.

Embodiment 2. FIG.
In the second embodiment, the human existence rate for each place / time is obtained by the moving body flow line creation unit 2 with ID and the moving body flow line creation unit 4, and the human existence rates are compared by the flow line comparison unit 5. The tag non-holder is detected. Since the configuration on the drawing is the same as that of the first embodiment shown in FIG. 1, description will be made with reference to FIG. The ID-attached mobile position detectors 1a, 1b, 1c and the mobile position detectors 3a, 3b, 3c in the second embodiment are the same as those in the first embodiment. The ID-attached moving body flow line creation unit 2 creates the tag space-time existence rate as the ID-attached moving body existence rate from the detection results of the ID-attached moving body position detection units 1a, 1b, and 1c. In addition, the moving body flow line creation unit 4 creates the space-time presence rate of the moving body as the moving body presence rate from the detection results of the moving body position detection units 3a, 3b, and 3c. Further, the flow line comparison unit 5 compares the moving object existence ratio with ID and the moving object existence ratio to obtain the existence ratio of the moving object having no tag. And the determination part 6 is comprised so that it may determine with the mobile body which does not have a tag existing, when the presence rate of the mobile body which does not have a tag is larger than predetermined value.

Next, the operation of the mobile management apparatus according to the second embodiment will be described.
First, create a flow line, and then determine the human presence rate.
In order to obtain the presence rate of the person with ID, the moving body flow line creation unit 2 with ID obtains a flow line for each tag ID. Since the flow line creation method is the same as that of the first embodiment, description thereof is omitted here. However, at the time of data integration in step ST202 of FIG. 2, instead of limiting the detection points to one, several candidate points may be selected. For each point, the reliability is calculated from the radio wave intensity and the like. When there are several candidate points, a plurality of flow lines are connected in step ST203 of FIG. Each flow line has the reliability obtained from the reliability of the candidate points. The reliability is normalized so as to be 1 when the reliability of the flow line at the same time is added. A method other than Embodiment 1 may be used as a method of creating a flow line.

Next, a method for calculating the human presence rate from the human flow with ID will be described. FIG. 10 is an explanatory diagram showing an example of obtaining the human presence rate from the ID-added human flow line in the second embodiment, and shows the human presence rate created from the flow line 402 of ID001 in FIG.
In practice, the tag reading position is not a point, but somewhere in the detection range area of the receiver. In addition, since the tag reading interval may be increased in order to extend the battery life of the tag, the reading timing is not always the moment when it enters the detection area. Therefore, when the existence probability in space-time is obtained based on the tag detection range characteristics of each receiver, for example, as shown in the graph 1001, the distribution is such that the probability decreases with increasing distance from the flow line with the flow line as the center. Become. The existence ratio of the tag ID 001 at time t has a distribution that becomes 1 when added in the entire space. For example, when the tag reception intensity of the receiver is approximated by a Gaussian distribution with a standard deviation σ, the human presence rate is a distribution obtained by moving the Gaussian distribution with the standard deviation σ along the flow line around the flow line. Since the human presence rate P (xp, t) at the time point t of the point P that is the position coordinate xp is a set of values on the straight line x = xp in the graph 1001, the graph 1002 is obtained.

  The distribution of the human presence rate was obtained from the detection characteristics of the receiver, but spatial attributes, personal walking characteristics, communication intensity at the time of tag ID reading, and the like may be used. For example, if the receiver (mobile body position detector with ID) 1a and the receiver (mobile body position detector with ID) 1b are installed in the passage and the walking direction of the passage is the direction of xa → xb, Since there is a low possibility that a person exists in the portion, the range of the human presence rate distribution in the direction orthogonal to the xa → xb direction can be limited to the width of the passage. For example, when the normal walking speed is known as the walking characteristic for each ID, the distribution of the presence rate is changed by comparing the actual moving speed calculated from the tag detection time with the normal walking speed. To do. For example, when the actual moving speed is faster than the normal walking speed, the probability of moving in a straight line is high, so the spread of the human presence rate distribution is reduced. Conversely, if the actual moving speed is slower than the normal walking speed, the possibility of meandering is high, and therefore the spread of the human presence rate is increased. When using the communication strength at the time of reading the tag ID, the distance between the receiver and the tag estimated from the communication strength is calculated using the characteristic that the distance between the receiver and the tag is closer as the communication strength is higher. Introduce into the distribution characteristics of abundance. The distribution becomes wider as the distance increases, and the distribution becomes narrower as the distance decreases. Thus, if the distribution of the human presence rate is calculated using the space attribute, the personal walking characteristics, the communication strength at the time of reading the tag ID, etc., the human presence rate can be obtained more accurately.

  When a plurality of flow lines exist at the same time, the human presence rate is obtained for each flow line, and the reliability of the flow line is added to the human presence rate. The presence rate obtained from all the flow lines is added to the human presence rate at the same time point. For example, at a certain time t, if there are n flow lines of the tag IDi and the existence rate of the point p (positional coordinate xp) obtained from the flow line n is Pin (xp, t), the point p of the tag IDi is The existence rate Pi (xp, t) of Pi becomes (Pi (xp, t) = reliability of the Σ flow line n * Pin (xp, t).

Next, a method for obtaining the human presence rate in the moving body flow line creation unit 4 will be described.
Since the method for creating a human flow line in the moving body flow line creation unit 4 is the same as that in the first embodiment, the description thereof is omitted here. However, at the time of data integration in step ST502 of FIG. 5, instead of narrowing the detection points to one point, several candidate points may be selected. 5 may be associated with a plurality of points, or a plurality of detection points associated with a plurality of correspondence points in step ST504 of FIG. 5 may be connected. Each flow line has the reliability obtained from the proximity of the detection points and the moving speed. A method other than Embodiment 1 may be used as a method of creating a flow line. Moreover, since the method of calculating | requiring a person presence rate from a human flow line is the same as the method of calculating | requiring the human presence rate for every ID from the flow line with ID mentioned above, the description is abbreviate | omitted.

  When the distance between the moving persons is sufficiently larger than the sensor installation interval and the installation range, it is possible to create a flow line from the detection pattern and obtain the distribution of the human presence rate. However, in places where people are crowded, it is difficult to divide the detection results of sensors that cannot identify individuals and connect them with a line, so there is a probability that a person exists around the time when the movement of the person is detected. The person presence rate may be calculated as high. At this time, the graph of the human presence rate is as shown by 1102 in FIG. The human presence rate is set around the time and position at which the human sensor detects the movement. A probability distribution function is used such that the time and position at which motion is detected has the highest presence rate, and the distribution gradually spreads in time and space. For example, it is assumed that a Gaussian distribution of standard deviation σΔt (Δt is the difference time from the detection time) with the detection position as the center is used, and σΔt is a function that increases with Δt. At points where human presence distributions obtained from a plurality of detection results overlap, by adding probabilities, a place where people's movements are dense in space and time has a higher probability of human presence. For example, since detection (xb, t6) and detection (xb, t7) are close in detection time, the human presence rate is added to the space between them.

Next, a method for comparing the human presence rate in the flow line comparison unit 5 will be described.
Assuming that the human presence rate at a point x at a certain time t is P (x, t) and the tag presence rate of IDi is Pi (x, t), the presence rate of the tag non-holder is P (x, t) −Pi. (X, t). The existence ratio of tag non-holders is obtained for all areas and all IDs, and when there are times / areas with a high presence ratio continuously in time and space, the determination unit 6 indicates that there are non-tag holders. To do. For example, detection of tag non-holders is performed by binarizing and clustering based on the existence probability of non-tag holders, and when a cluster having a high existence probability is larger than a certain level, there is a non-tag holder in the determination unit 6. Suppose that

As described above, in the mobile management apparatus according to the second embodiment, the flow line of the person is extracted from the detection pattern in order to detect the tag non-retainer by the presence rate of the person and the tag without uniquely specifying the flow line. This embodiment can be applied even when it is difficult or when ambiguity remains in the flow line.
In addition, in the sensor arrangement management of the ID-attached moving body position detecting units 1a, 1b, 1c and the moving body position detecting units 3a, 3b, 3c, only the coordinate information indicating where each sensor is arranged is provided. Since the human presence rate can be calculated, when changing the sensor position, it is only necessary to change the setting of the sensor whose position has been changed, and it is not necessary to change the setting of the sensor whose position has not been changed. If you have an association table between sensors, if you change the position of one sensor, you must also change the settings of other sensors that did not move, which may increase work costs or cause a change error. .
Furthermore, if the distribution of the human presence rate is calculated using spatial attributes, personal walking characteristics, communication intensity at the time of reading the tag ID, etc., the human presence rate can be obtained more accurately.

  As described above, according to the mobile unit management apparatus of the second embodiment, the mobile unit management apparatus identifies a mobile unit having a tag and a mobile unit having no tag, and detects the position of the tag. Mobile unit with ID for creating the presence rate of the space-time of the tag as the mobile unit presence rate with ID from the detection results of the mobile unit position detection units with the ID 1a, 1b, 1c and the ID-attached mobile unit position detection units 1a, 1b, 1c From the detection results of the flow line creation unit 2, the mobile body position detection units 3a, 3b, and 3c that detect the mobile body, and the mobile body position detection units 3a, 3b, and 3c, the existence rate of the time and space of the mobile body is determined. The moving body flow line creation unit 4 that creates the rate, the moving body existence rate with ID and the moving body existence rate are compared, the flow line comparison unit 5 that obtains the existence rate of the moving body that does not have a tag, and a tag If the presence rate of moving objects that are not Since there is a determination unit 6 that determines that there is no moving body, it does not have a tag even if it is difficult to extract a person's flow line, or even if ambiguity remains in the flow line The moving object can be accurately determined.

Embodiment 3 FIG.
In the third embodiment, the ID-added moving body flow line creation unit 2 and the moving body flow line creation unit 4 obtain the number of people distribution for each place and time, and the flow line comparison unit 5 compares the number of people to each tag. It is designed to detect non-holders. Since the configuration on the drawing is the same as that of the first embodiment shown in FIG. 1, description will be made with reference to FIG. The moving body position detection units 1a, 1b, 1c with ID and the moving body position detection units 3a, 3b, 3c in the third embodiment are the same as those in the first embodiment. The ID-attached moving body flow line creating unit 2 creates the number distribution of tags in space and time as the number of ID-added person distribution from the detection results of the ID-attached moving body position detecting units 1a, 1b, and 1c. In addition, the moving body flow line creation unit 4 creates a time and space distribution of the moving body from the detection results of the moving body position detection units 3a, 3b, and 3c. Furthermore, the flow line comparison unit 5 compares the number distribution with ID and the number distribution to obtain the number distribution of persons who do not have a tag. And the determination part 6 is comprised so that it may determine with the person who does not have a tag existing, when the number distribution of the person who does not have a tag is more than predetermined value.

Hereinafter, the operation of the mobile unit management apparatus according to the third embodiment will be described.
The operations of the ID-attached mobile body position detection units 1a, 1b, and 1c are the same as those in the first and second embodiments. The ID-attached moving body flow line creation unit 2 calculates the number distribution by adding the presence rates of all IDs. In the moving body flow line creation unit 4, for example, the number distribution is estimated from the number of feature points shown in the camera image and the detection frequency of the human sensor in the area. The true value of the total number of people on the floor or building may be obtained from the entrance / exit management device, the distribution of people may be calculated based on the detection frequency, and the total number of people may be allocated based on the distribution of people. For example, if there are two areas A and B on the floor, the detection frequency for each area is A: B = 8: 2, and the total number of floors is 10, the area A has 8 people and the area B has Suppose there are two people. In the flow line comparison unit 5, the difference between the number distribution obtained by the moving body flow line creation unit 4 and the number distribution with ID obtained by the moving body flow line creation unit 2 with ID is obtained for all areas and all IDs. The distribution of the number of non-tag holders. The determination unit 6 determines that there is a tag non-holder when the number distribution of tag non-holders is equal to or greater than a predetermined value.

  As described above, in the third embodiment, the tag non-retainer is detected using the distribution of the number of people rather than the mere presence rate of the person. Therefore, even when the tag holder and the non-tag holder are acting together, Can detect non-holders. In the local comparison of the number of persons, erroneous detection of a tag non-holder may occur, but accuracy is improved because long-term tracking is performed to detect a tag non-holder.

  As described above, according to the mobile unit management apparatus of the third embodiment, the mobile unit management apparatus identifies a mobile unit having a tag and a mobile unit having no tag, and detects the position of the tag. ID-attached movement for obtaining the moving object number distribution in the space-time of the tag as the ID-attached moving object number distribution from the detection results of the attached object position detecting parts 1a, 1b, 1c and the ID-attached moving object position detecting parts 1a, 1b, 1c From the detection results of the body movement line creation unit 2, the mobile body position detection units 3a, 3b, and 3c that detect the mobile body, and the mobile body position detection units 3a, 3b, and 3c, the moving body number distribution in the time and space of the mobile body is obtained. The moving body flow line creation unit 4 to be obtained, the moving body number distribution with ID and the moving body number distribution are compared, the flow line comparison unit 5 to obtain the moving body number distribution of the moving body having no tag, and no tag. If the moving object number distribution of moving objects is greater than or equal to a predetermined value, tag Because the determination unit 6 that determines that there is a moving body that does not exist is provided, for example, even when a tag holder and a tag non-holder carry out actions together, the tag non-holder is accurately detected. Can do.

Embodiment 4 FIG.
In the fourth embodiment, a learning period for a true value and measurement data is provided before actual operation, and the system automatically adjusts a threshold value and the like. Since the basic configuration of the ID-attached mobile body position detection units 1a, 1b, and 1c to the determination unit 6 is any one of the first to third embodiments, detailed description thereof is omitted here.

  In the fourth embodiment, machine learning may be used when calculating the movement line / person presence rate / number distribution in the ID-attached moving body flow line creation unit 2 or the moving body flow line creation unit 4. For example, several patterns are learned by combining the flow line, the true value of the presence / absence / number of persons, and the detection patterns of the ID-attached mobile body position detectors 1a, 1b, 1c and the mobile body position detectors 3a, 3b, 3c. In addition, the flow line, the presence / absence of persons, the number of persons, etc. are estimated according to which of the patterns learned by the ID-attached mobile body position detectors 1a, 1b, 1c and the mobile body position detectors 3a, 3b, 3c. Moreover, the walking speed of the tag holder is automatically learned from the position of the receiver that has read the tag ID and the detection time, and is reflected when creating the flow line in the ID-attached moving body flow line creating unit 2 and the moving body flow line creating unit 4 You may let them. When the flow line comparison unit 5 compares the flow lines, the moving body flow line creation unit 2 with ID and the mobile body flow line creation unit 4 compare the flow lines of the tag holders in advance. The difference in the flow line at the time of detection may be used as the threshold value of the flow line comparison unit 5.

  As described above, according to the moving body management apparatus of the fourth embodiment, at least one processing unit among the ID-attached moving body flow line creation unit 2, the mobile body flow line creation unit 4, and the flow line comparison unit 5. Since learning of the processing result is performed based on the true value and measurement data prepared in advance, the accuracy of the flow line creation and comparison can be improved. In addition, errors due to the installation environment can be absorbed by learning, improving accuracy. Furthermore, parameter adjustment at the time of installation can be automated, and the labor of the installer can be saved.

Embodiment 5 FIG.
In the fifth embodiment, the ID-attached mobile body motion is based on the positional relationship of the installation positions of the detectors in the ID-attached mobile body position detectors 1a, 1b, and 1c and the mobile body position detectors 3a, 3b, and 3c. The line creation unit 2 and the moving body flow line creation unit 4 each create a flow line or the like. Since the configuration on the drawing is the same as that of the first embodiment, description will be made with reference to FIG.

  In addition to managing the installation position of the tag receiver and human sensor by coordinates, the movement in the ID-attached moving body flow line creation unit 2 and the moving body flow line creation unit 4 can be performed by using the adjacent relationship between devices. Line creation accuracy is improved. For example, when tag receivers (mobile unit position detectors with ID) 1a, 1b, and 1c are installed in the passage as shown in FIG. 12, if there is a tag detection error in the receiver 1b, The line creation unit 2 may connect the flow lines of the receiver 1a and the receiver 1c. Therefore, the fifth embodiment uses the adjacency relationship between devices, that is, the relationship that the receiver 1a and the receiver 1b are adjacent to each other and the receiver 1b and the receiver 1c are adjacent to each other. Thereby, it can be estimated that the receiver 1b is always passed between the detection of the receiver 1a and the receiver 1c, and the flow line can be correctly connected. When spatial information such as walls and passages are stored as information on the same coordinate system as the installation position of the receiver, the collision between the flow line between the receiver 1a and the receiver 1c and the wall is geometrical. Can be calculated, but the calculation cost is high.

  Also, the setting of the arrangement position of the device, the non-intrusive area, and the adjacent relationship between the devices may be managed as a table, or may be set on the GUI and the management table may be automatically generated. For example, the installation position coordinates of the device are set by installing a device icon on the floor drawing. Similarly, an inviolable area is set on the drawing. Here, the device adjacency information may be automatically generated from the human inviolable area and the position coordinates of the device installation.

  As described above, according to the mobile management apparatus of the fifth embodiment, the ID-attached mobile body flow line creation unit 2 and the mobile body flow line creation unit 4 each include a plurality of ID-attached mobile body position detection units 1a and 1b. , 1c and the plurality of moving body position detectors 3a, 3b, 3c, the processing is performed based on the positional relationship of the installation positions of the detectors, so that calculation with high calculation costs such as collision determination and neighborhood search is performed. Therefore, since it is possible to obtain the position of a nearby device, the connection of the space, and the like, the calculation cost can be reduced.

Embodiment 6 FIG.
In the sixth embodiment, the flow line comparison unit 5 performs multi-stage comparison.
When the moving body flow line creation unit 2 with ID and the moving body flow line creation unit 4 obtain the flow line and existence rate for each point and compare with the flow line comparison unit 5, the calculation cost becomes enormous. The calculation and the position comparison may be performed for each space interval and time interval, and the granularity (the interval between the space interval and the time interval) may be dynamically changed. The time line and space section to be compared by the flow line comparison unit 5 are compared widely, and the time section and space section where there is a high possibility that a tag non-holder is present are divided again and divided for comparison. That is, in the sixth embodiment, the flow line comparison unit 5 first performs comparison processing for each predetermined space section or time section as the first processing, and then performs the first processing space section or time section. The second processing is performed by narrowing the interval and limiting the target spatial region or time region. Note that the second process may be performed not only once but a plurality of times. Further, the time interval used for the comparison may be changed according to the possibility that there is a tag non-holder. For each space section, the time interval used for comparison may be narrowed when the possibility of having a tag non-holder is high, and the time interval used for comparison may be lengthened when the possibility of having a non-tag holder is low.

  As described above, according to the mobile management apparatus of the sixth embodiment, the comparison process in the flow line comparison unit 5 includes the first process performed for each predetermined space section or time section, and the space of the first process. Since the interval or time interval is narrowed and the second process is limited to the target spatial region or time region, the temporal and spatial intervals for comparison can be set in multiple steps or dynamically. By changing, the calculation cost can be reduced without reducing accuracy.

Embodiment 7 FIG.
When detecting non-tag holders for monitoring purposes, it is desirable to detect non-tag holders as soon as possible, but to create and compare the flow lines of all people in places with a large number of people such as office buildings. Is very computationally expensive. Therefore, in the mobile management device of the seventh embodiment, for example, the assigned area section is divided and calculations are performed in parallel.

FIG. 13 is a configuration diagram illustrating the mobile management apparatus according to the seventh embodiment.
In the figure, the ID-attached mobile body position detection units 1a, 1b, and 1c and the mobile body position detection units 3a, 3b, and 3c are the same as those in the first embodiment. A plurality (n) of moving body flow line creation units 2-1 to 2-n with ID, moving body flow line creation units 4-1 to 4-n, and flow line comparison units 5-1 to 5-n are provided. Each of them is an arithmetic processing unit that performs processing in parallel. The processing content is the same as that of the ID-attached moving body flow line creation unit 2, the moving body flow line creation unit 4, and the flow line comparison unit 5 of any embodiment. The determination unit 6a is configured to perform determination processing using the results of the plurality of flow line comparison units 5-1 to 5-n.

  In the mobile management apparatus according to the seventh embodiment, for example, the assigned area section is divided and calculations are performed in parallel. In addition, about this division | segmentation method, the mobile body position detection part 1a, 1b, 1c with ID and the mobile body position detection part 3a, 3b, 3c may be sufficient, and another division | segmentation method may be sufficient. The determination unit 6a detects a tag non-holder by using the results of the plurality of flow line comparison units 5-1 to 5-n.

  In the above description, there are a plurality of ID-attached mobile body flow line creation units 2-1 to 2-n, mobile body flow line creation units 4-1 to 4-n, and flow line comparison units 5-1 to 5-n. Although the case where it is provided has been described, only one of the arithmetic processing units may be processed in parallel, for example, such that only the flow line comparison units 5-1 to 5-n are plural.

Further, the comparison history in the flow line comparison unit 5 may be separately stored, and determination may be made using this comparison history, which will be described next.
FIG. 14 is a configuration diagram illustrating another example of the mobile management apparatus according to the seventh embodiment.
The mobile body management device shown in FIG. 14 includes ID-attached mobile body position detection units 1a, 1b, and 1c, ID-attached mobile body flow line creation unit 2, mobile body position detection units 3a, 3b, and 3c, and mobile body flow line creation unit. 4, a flow line comparison unit 5a, a determination unit 6b, and a comparison history storage unit 7 are provided. Here, the ID-attached moving body position detection units 1a, 1b, and 1c to the moving body flow line creation unit 4 are the same as those in any one of the first to sixth embodiments. The flow line comparison unit 5a is configured to compare the flow lines newly multiplied by the ID-attached moving body flow line creation unit 2 and the moving body flow line creation unit 4 and store the comparison history in the comparison history storage unit 7. . The determination unit 6b is configured to determine detection of a tag non-holder based on the comparison history in the comparison history storage unit 7 and the result of the flow line comparison unit 5a.

  In the mobile management apparatus configured as described above, the flow line comparison unit 5a only compares the flow lines newly generated in the ID-attached mobile flow line creation unit 2 and the mobile body flow line creation unit 4, and the comparison history is Stored in the comparison history storage unit 7. And the determination part 6b determines the detection of a tag non-holding person from the comparison history memorize | stored in the comparison history memory | storage part 7, and the result of the flow line comparison part 5a.

  As described above, according to the mobile management apparatus of the seventh embodiment, at least one of the processing units including the ID-attached mobile flow line creation unit 2, the mobile body flow line creation unit 4, and the flow line comparison unit 5 is provided. Since a plurality of processing units are provided and the processing units provided in parallel perform processing in parallel, it is possible to quickly detect a mobile object that does not have a tag.

  In addition, according to the mobile body management device of the seventh embodiment, the mobile unit management apparatus includes the comparison history storage unit 7 that stores the history of comparison results in the flow line comparison unit, and the determination unit 6b includes the comparison result of the flow line comparison unit 5a, Since the determination is made based on the history of comparison results stored in the comparison history storage unit 7, a plurality of ID-attached mobile body flow line creation units 2, mobile body flow line creation units 4 and flow line comparison units 5 are provided. In addition, a mobile object having no tag can be detected in real time.

Embodiment 8 FIG.
In the eighth embodiment, the type of flow line created by the ID-attached mobile flow line creation unit 2 and the mobile flow line creation unit 4 and the comparison method by the flow line comparison unit 5 are changed for each space or time zone. It is what I did. Since the configuration on the drawing is the same as that of the first embodiment, description will be made with reference to FIG.
The ID-attached moving body flow line creation unit 2 according to the eighth embodiment creates a time-space flow rate or presence rate of the tag based on the detection results of the ID-attached moving body position detection units 1a, 1b, and 1c. To do. In addition, the moving body flow line creation unit 4 creates a time line or a presence rate or a number distribution of the moving body from the detection results of the moving body position detection units 3a, 3b, and 3c. That is, the ID-attached moving body flow line generating unit 2 and the moving body flow line generating unit 4 are the ID-attached moving body flow line generating unit 2 and the moving body flow line generating unit 4 of the first, second, and third embodiments described above. One of these functions. The flow line comparing unit 5 selects and compares the results of the ID-attached moving body flow line creating unit 2 and the moving body flow line creating unit 4 based on spatial factors or temporal factors. Alternatively, the processing is performed by changing the space section or time section of the flow line comparison based on a spatial factor or a temporal factor. The determination unit 6 determines a moving body that does not hold a tag based on the comparison result of the flow line comparison unit 5.

  For example, in a place where there are many tag holders or in the vicinity of a tag of a monitor, since monitoring is performed by human eyes, the space interval and the time interval of the flow line comparison are sparse. In addition, in areas where people are crowded, such as conference room areas, the reliability of the flow line created by the moving body flow line creation unit 4 is low, so tags are not retained using the human presence rate instead of the flow line In a space where people are sparse, such as a passage, a person who does not hold a tag is detected using a flow line. The type of data used for the flow line comparison may be determined for each space in advance, or based on the flow lines and human distribution information created by the ID-attached mobile body flow line creation unit 2 and the mobile body flow line creation unit 4 The comparison method in the flow line comparison unit 5 may be changed.

Further, in the eighth embodiment, as shown in FIG. 15, the flow line and human distribution information created by the ID-attached moving body flow line creating unit 2 and the moving body flow line creating unit 4 and the flow line comparing unit 5 Detection parameters such as sensor sensitivity in the ID-attached mobile body position detection units 1a, 1b, and 1c and the mobile body position detection units 3a, 3b, and 3c may be dynamically changed using any of the comparison results. In FIG. 15, among the plurality of ID-attached mobile body position detectors 1a, 1b, 1c and mobile body position detectors 3a, 3b, 3c, the ID-attached mobile body position detector 1a and mobile body position detector 3a. Only shows. In addition, the ID-attached mobile body position detection units 1a, 1b, 1c and the determination unit 6 other than changing detection parameters of the mobile body position detection units 3a, 3b, 3c are used. The basic configuration is the same as that of any one of the first to sixth embodiments.
The detection parameter is changed, for example, by increasing the detection sensitivity of the human sensor in an area where human movement is small, such as during a meeting. From the history of the direction of the flow line, the trend of the flow line direction in the space is obtained, and this is used when creating the flow line or obtaining the human presence rate from the flow line.

  As described above, according to the moving body management apparatus of the eighth embodiment, the moving body management apparatus identifies a moving body having a tag and a moving body having no tag, and detects the position of the tag. From the detection results of the attached mobile body position detection units 1a, 1b, and 1c and the ID-attached mobile body position detection units 1a, 1b, and 1c, any one of the flow line, the existence rate, and the mobile body number distribution of the time space of the tag ID-attached moving body flow line creation unit 2 selected and created based on spatial factors or temporal factors, moving body position detection units 3a, 3b, 3c for detecting a moving body, and moving body position detection units 3a, 3b , 3c, a moving body flow line creation unit 4 for selecting and creating any one of a moving line in time and space, a presence rate, and a moving body number distribution based on a spatial factor or a temporal factor from the detection result of 3c, , Moving body flow line creation unit 2 with ID and moving body flow line creation unit 4 Since the flow line comparison unit 5 that compares the results and the determination unit 6 that determines the moving body that does not hold the tag based on the comparison result of the flow line comparison unit 5 are provided, it is matched to the spatial attribute and the movement of the moving body. Thus, an optimal comparison method can be selected, and a mobile object having no tag can be accurately determined.

  In addition, according to the moving body management apparatus of the eighth embodiment, the spatial section or the time section of the flow line comparison based on the results of the ID-attached moving body flow line creating unit 2 and the moving body flow line creating unit 4 is determined as a spatial factor. Or since it changed based on a time factor, optimal comparison can be performed according to a spatial attribute or a motion of a mobile body, and the mobile body which does not have a tag can be determined exactly.

  In addition, according to the moving body management apparatus of the eighth embodiment, the calculation result of at least one of the processing units among the ID-attached moving body flow line creation unit 2, the mobile body flow line creation unit 4, and the flow line comparison unit 5. Based on the above, the detection parameters of at least one of the ID-attached mobile body position detectors 1a, 1b, and 1c and the mobile body position detectors 3a, 3b, and 3c are changed. Optimal position detection can be performed according to the movement of the body, and a mobile body having no tag can be accurately determined.

  1a, 1b, 1c ID-attached mobile body position detector, 2, 2-1 to 2-n ID-attached mobile body flow line generator, 3a, 3b, 3c Mobile body position detector, 4, 4-1 to 4- n moving body flow line creation unit, 5, 5a, 5-1 to 5-n flow line comparison unit, 6, 6a, 6b determination unit, 7 comparison history storage unit.

Claims (11)

A mobile management device for identifying a mobile body having a tag and a mobile body having no tag,
An ID-attached mobile position detector that detects the position of the tag;
An ID-attached moving body flow line creation unit that creates a time-space flow line of the tag as an ID-attached moving body flow line from the detection result of the ID-attached moving body position detection unit;
A moving body position detector for detecting the moving body;
A moving body flow line creating unit that creates a moving line in the time and space of the moving body as a moving body flow line from the detection result of the moving body position detecting unit;
A flow line comparison unit that compares the moving body flow line with ID and the mobile body flow line, and associates the movement line close to the movement line;
A moving body management apparatus comprising: a determining section that determines, as the moving body having no tag, a moving body having a flow line that is not associated in the flow line comparing unit. A mobile management device for identifying a mobile body having a tag and a mobile body having no tag,
An ID-attached mobile position detector that detects the position of the tag;
An ID-attached mobile body flow line creation unit that creates the presence rate of the space-time of the tag as an ID-attached mobile body presence rate from the detection result of the ID-attached mobile body position detection unit;
A moving body position detector for detecting the moving body;
A moving body flow line creation unit that creates a moving object presence rate as a moving object existence rate from the detection result of the moving object position detection unit;
A flow line comparison unit that compares the mobile body presence rate with ID and the mobile body presence rate and obtains the mobile body presence rate without the tag;
A moving body management apparatus comprising: a determination unit that determines that a moving body that does not have a tag is present when a presence rate of the moving body that does not have the tag is greater than or equal to a predetermined value. A mobile management device for identifying a mobile body having a tag and a mobile body having no tag,
An ID-attached mobile position detector that detects the position of the tag;
An ID-attached moving body flow line creation unit that obtains a moving object number distribution in the spatio-temporal space of the tag as an ID-attached moving body number distribution from the detection result of the ID-attached moving body position detection unit;
A moving body position detector for detecting the moving body;
A moving body flow line creation unit for obtaining a moving body number distribution in space-time of the moving body from the detection result of the moving body position detection unit;
A flow line comparison unit for comparing the ID-equipped mobile body number distribution with the mobile body number distribution and obtaining a mobile body number distribution of the mobile body not having the tag;
A moving body management apparatus comprising: a determination unit that determines that a moving body that does not have a tag exists when a moving body number distribution of the moving body that does not have the tag is equal to or greater than a predetermined value.   At least one of the moving body flow line creation unit with ID, the mobile body flow line creation unit, and the flow line comparison unit learns the processing result based on the true value and measurement data prepared in advance. The mobile object management apparatus according to claim 1, wherein   The ID-attached mobile body flow line creation unit and the mobile body flow line creation unit perform processing based on the positional relationship of the installation positions of the detection units in the plurality of ID-attached mobile body position detection units and the plurality of mobile body position detection units, respectively. The mobile body management device according to claim 1, wherein the mobile body management device is performed.   The comparison process in the flow line comparison unit includes a first process performed for each predetermined space section or time section, and a space section or time section of the first process is narrowed, and a target space area or The mobile management apparatus according to any one of claims 1 to 5, characterized by comprising a second process in which a time region is limited.   There is provided a plurality of at least one processing unit among an ID-attached mobile body flow line creation unit, a mobile body flow line creation unit, and a flow line comparison unit, and the plurality of processing units perform processing in parallel. The mobile body management device according to any one of claims 1 to 6. A comparison history storage unit for storing a history of comparison results in the flow line comparison unit;
7. The determination unit according to claim 1, wherein the determination unit performs determination based on a comparison result of the flow line comparison unit and a history of comparison results stored in the comparison history storage unit. The mobile management apparatus according to 1. A mobile management device for identifying a mobile body having a tag and a mobile body having no tag,
An ID-attached mobile position detector that detects the position of the tag;
With ID for selecting and creating one of the temporal and temporal flow lines of the tag, the presence rate, and the moving body number distribution from the detection result of the ID-attached moving body position detecting unit based on the spatial factor or the temporal factor A moving body flow line creation unit;
A moving body position detector for detecting the moving body;
A mobile body motion that selects and creates any one of the temporal and temporal flow lines, the presence rate, and the mobile body number distribution of the mobile body from the detection result of the mobile body position detection unit based on a spatial factor or a temporal factor. A line creation section;
A flow line comparison unit for comparing the results of the moving body flow line creation unit with ID and the moving body flow line creation unit;
A moving body management apparatus comprising: a determination unit that determines a moving body that does not hold the tag based on a comparison result of the flow line comparison unit.   The space section or the time section of the flow line comparison based on the results of the ID-attached moving body flow line creation unit and the moving body flow line creation unit is changed based on a spatial factor or a temporal factor. The mobile body management device according to claim 9.   Based on the calculation results of at least one of the processing units including the ID-attached moving body flow line creation unit, the mobile body flow line creation unit, and the flow line comparison unit, the ID-attached mobile body position detection unit and the mobile body position detection unit The mobile body management apparatus according to any one of claims 1 to 10, wherein a detection parameter of at least one of the detection units is changed.

Images