JP2004096501A - System and method for detecting position of mobile object and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a system for detecting the position of a mobile object, a system which individually and continuously detects the position of a child (infant) and with which the child is prevented from being continuously exposed to an electromagnetic wave like in a case when PHS(R) is used, and to provide a method therefor. <P>SOLUTION: A traveling rate detection means 10 such as an acceleration sensor is mounted on each infant 100 nursed in a nursery and the action of the infant is continuously recorded. Further, the infant is photographed with a position detection means 11 such as a TV camera for picture processing. The position of each infant 100 is detected without identifying the infant itself. The traveling rate information of the infant 100 which the traveling rate detection means 10 detects is collated with positional information of the infant 100 which is detected by the position detection means 11, and the continuous position of each infant 100 is detected by specifying each infant. <P>COPYRIGHT: (C)2004,JPO

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention, the position of a mobile object, for example, an infant or the like in a nursery school, there is a possibility of damaging the human body such as electromagnetic waves (radio waves), or without using physical signal means that cannot deny that there is a danger. The present invention relates to a moving object position detection system and a moving object position detection method that can be specified continuously from the time of opening the park to the time of closing the park.
[0002]
[Prior art]
In recent years, due to the involvement of parents in society, infants are often unable to spend their days in their parents'"hugging". From these points, the role that preschool educational institutions such as nursery schools and kindergartens should play is becoming more important.
[0003]
For infants who spend their days in kindergartens and nursery schools, it is an important issue to monitor the activities of each individual as much as possible and raise them under a childcare plan that matches each individual's individuality.
[0004]
On the other hand, in nursery schools and kindergartens, it is experts who have received professional training at universities such as nursery teachers or kindergarten teachers and have acquired national qualifications. However, there may be many inexperienced nursery teachers and kindergarten teachers.
[0005]
From the above points, it is important from the viewpoint of child care support to introduce various IT (information technology) into kindergartens and nursery schools and reduce the burden on inexperienced nursery teachers and kindergarten teachers. As one of such approaches, one possibility is to detect and record the position of children in a nursery school or kindergarten.
[0006]
For example, nursery teachers and kindergarten teachers can tell parents who have been playing with each other today. It is natural for parents to want to know what's going on in their children's garden. However, nursery teachers and kindergarten teachers who have many children as homeroom teachers have difficulty responding to all children. If the positional information is known for each infant, it is possible to detect who and who are acting together.
[0007]
In general, whether a moving object is a living thing such as a human or an inanimate object, it is not easy to detect a position with high accuracy, small size, and easy in a large area. At present, a representative example of a method experimentally performed is position detection using PHS (registered trademark) (Personal Handyphone System).
In the case of PHS (registered trademark), the position of the PHS (registered trademark) terminal is controlled to a certain degree by a method such as three-point surveying by receiving radio waves of one PHS (registered trademark) terminal by a plurality of base stations. I understand. Therefore, if the infant is provided with the PHS (registered trademark), the position of each infant can be determined. Normally, PHS (registered trademark) ground stations are installed at a distance of about 100 meters, but they must be placed close to each other on the premises or rooms of nursery schools and kindergartens, from several tens of meters to about 10 meters. It is known that the position accuracy can be increased and the position can be detected with an error of 1 meter or less.
[0008]
However, this method has a major problem. This is a problem of electromagnetic waves (radio waves). Needless to say, in the "small zone system" wireless system such as PHS (registered trademark), a terminal transmits a radio wave for notifying its own position to the base station only when the terminal moves and enters a different small zone. I do. Also, the electric field strength of the transmission output radio wave of PHS (registered trademark) is smaller than that of a mobile phone. However, if the position of the terminal is detected only at the moment when the small zone is switched, it is inappropriate for the position detection of the infant. In other words, the position of the infant must be reported to the nearby base station by transmitting radio waves frequently, for example, every one minute or every three minutes. This frequency, that is, the accumulated power of the transmitted radio wave, increases as the position measurement accuracy is increased. Even if PHS (registered trademark) has low air power, continuing to emit radio waves all the time may cause anxiety for parents of infants due to radio wave exposure.
[0009]
Another major problem is that position detection by PHS (registered trademark) has poor resolution. For example, suppose you are looking at a picture-story show. It is one of the observation items in nursery teacher guidance or in research on the method of early childhood education to see whether they are actively moving their hands or feet in response to a picture-story show scene change. However, such a thing cannot be preserved unless everyone's image is also captured in video. Moreover, it cannot be analyzed automatically by a computer using image processing. At present, research on early childhood education methodology is being conducted in such a way that one observer is assigned to one infant. On the other hand, even if the behavior of the infant is chased by the PHS (registered trademark) (ignoring the influence of the radio), even the movement of the body cannot be identified with the rough resolution of the PHS (registered trademark). That is, position detection by PHS (registered trademark) is a position detection method with low accuracy, but which may involve the danger of radio wave exposure.
[0010]
On the other hand, as a method of detecting a human motion, for example, there is an image processing method. In the image processing method, a human portion is cut out from an image captured by a camera or the like by image processing. To facilitate the detection, a reflector having specific optical properties may be attached to the human body, or a personal identification means such as a barcode may be attached to the human body. This is an easy way to recall. The main feature of this method is that it is non-invasive, utilizing the visible light that has already been hit normally. For example, research has been widely known in which a camera is photographed from the ceiling in an elevator to detect the position of a person and count the number of people.
[0011]
The image processing method is friendly to the human body because it does not use radio waves, ultrasonic waves, infrared rays, or the like, which may affect the human body. On the other hand, it has some problems. First of all, not much accuracy can be expected. There is a danger of counting two people who are shouldering each other as one, or counting a cat as a human. These are the basic properties of the pattern recognition device. In pattern recognition, tolerance for erroneous recognition is always required. Further, the image processing method cannot detect fine body movements similarly to PHS (registered trademark). Therefore, only rough movements can be grasped. This is the same as PHS (registered trademark).
[0012]
Further, as a decisive problem, in the image processing method, it is not possible to identify which of a plurality of humans and infants is photographed unless a bar code or the like is attached to each child. Of course, it is conceivable in principle that each child is given a different barcode or color code, and personal identification is thereby performed. However, this may also be covered with a hand or the like, and strict restrictions may be added to lighting (illumination) conditions with a color code or the like. Further, in order to read the barcode of each of the children scattered in a large room, a camera having an enormous resolution must be used, and a large barcode needs to be attached to the child. This is also unrealistic.
[0013]
As described above, in the image processing method, although there is no danger to the human body due to the radio waves possessed by PHS (registered trademark), the obtained data is too coarse as in PHS (registered trademark). In addition, it is inferior to PHS (registered trademark) in that there is a possibility of erroneous recognition. In the above description, an image of detecting a position based on a captured image of a camera or the like has been described. However, non-invasive position detection goes beyond this. For example, a pressure sensor may be attached to the floor. Further, instead of a normal camera, a system that detects infrared rays may be used. Here, the position of an infant or the like can be detected to the last, but it is a disadvantage of many non-invasive position detection methods that it is not known who the child is and that there is a detection error.
[0014]
On the other hand, although the position cannot be specified accurately, there is a method of collecting personal information by identifying the person from the beginning. Originally, when a sensor is attached to the individual, it is obvious that the individual identification has been performed from the beginning. As a method for acquiring such individual movement, there is a method using a sensor for measuring acceleration or angular acceleration. The measurement itself may be based on a sensor worn on the human body, and it is not necessary to apply any additional physical signals to the human body. Instead of acceleration, a sensor for velocity or angular velocity may be assumed. For example, as a speed sensor, an approach that detects the movement of air around the body can be considered. What kind of measuring means is used is not limited in the present invention. In the present invention, the use of an “acceleration sensor” is considered as an example. When the acceleration sensor is attached to the infant, the small body movement of the infant can be detected. Already, instruments for calculating the calorie consumption of humans or counting the number of steps from acceleration or angular acceleration are widely available on the market. In these devices, target data such as calories are obtained in consideration of parameters such as weight, height, and gender based on measured values within a certain period of time. The ability to deal with such fine data is the first merit of the method using an acceleration sensor or the like.
[0015]
However, there is a major problem in using an acceleration sensor or the like. That is, the frequency range that can be measured is limited. Therefore, even if the raw output of the acceleration sensor is integrated twice, the position of the infant cannot be accurately calculated. In particular, errors accumulate and become fatal as the time elapses. This is a problem inherent to the sensor because the DC component drifts due to temperature or the like. Therefore, the acceleration sensor alone does not provide information that is useful for child-rearing support, such as where in the room or with whom (even if calorie consumption and steps are known).
[0016]
[Problems to be solved by the invention]
As described above, without using any communication means such as radio waves, ultrasonic waves, and infrared rays that affect the human body at all, the position of a moving object such as an infant in a room or a garden is detected, and at the same time, Also, it is necessary to measure the movement of the infant.
[0017]
In the present invention, position measurement means by image recognition, pressure detection, or the like, in which an individual cannot be identified (here, “unspecified” means that a child (human) can be detected there, "Cannot recognize who the person is)", and effectively combine the features of each of the movement amount measuring means such as an acceleration sensor capable of detecting fine movements of the individual, and determine the position of the child (infant). And a method for detecting the position of a moving object, in which a child is not continuously exposed to electromagnetic waves as in the case of using PHS (registered trademark) while detecting the position of the moving object individually and continuously. The purpose is to do.
[0018]
[Means for Solving the Problems]
The present invention has been made to solve the above problems, and is a position detection system of a moving body for detecting a continuous position of one or more moving bodies, the system being attached to each of the moving bodies. Moving amount detecting means for continuously detecting the moving amount of the moving object, and individually detecting the position of each moving object continuously or discontinuously without identifying the moving object itself. Position detecting means, and the moving amount information detected by the moving amount detecting means, and the position information of the moving object detected by the position detecting means are compared with each other, and the successive positions of the moving objects are determined by the moving object. And a matching means for detecting while identifying the information.
Thereby, the positions of a plurality of children can be individually and continuously detected. Further, unlike the case of position detection using PHS (registered trademark), a child does not continuously receive radio waves. Also, for example, even when a plurality of children enter a tunnel (for a game) and do not know who came out first, it is possible to correctly grasp the position information and identify the individual. Further, for example, if the movement amount data is taken out to a server or the like at the entrance of the door, the position of the child until the child passes through the door can be traced from at least the movement amount up to that point and the position measurement result by the TV camera. .
[0019]
Also, the moving object position detection system of the present invention uses an acceleration, speed, angular acceleration, or angular velocity sensor as the movement amount detection means, and uses an image processing camera or a weight sensor as the position detection means. It is characterized by the following.
Thus, for example, a sensor capable of detecting a fine motion, such as an acceleration gyro, can be attached to an infant to obtain data that can be used for a fine evaluation such as the effect of infant education. Further, in position detection by PHS (registered trademark), the more the accuracy of the position is to be improved, the shorter the interval at which radio waves are transmitted, and the more the infant is exposed to radio waves. However, in the present invention, since the position is detected by the TV camera or the like, extremely fine measurement of about 1/30 second can be performed from the beginning.
[0020]
Further, the moving object position detection system of the present invention is a moving object position detecting system for detecting a continuous position of one or more moving objects, and is attached to each of the moving objects. Moving amount detecting means for continuously detecting the moving amount of the moving object, and the position of each moving object can be individually detected continuously or discontinuously without identifying the moving object itself. The moving amount information detected by the moving amount detecting unit and the position information of the moving object detected by the position detecting unit to identify the continuous position of each moving object, The position detection in the time period in which the position detecting means cannot continuously detect the position of the moving object can be detected in the time period before and after the moving amount detecting means and the position detecting means. Based on the collation relationship , Characterized by comprising a reference means for detecting the position of the moving object only by the output movement amount information of the moving amount detecting means.
Thereby, the positions of a plurality of children can be individually and continuously detected. Further, unlike the case of position detection using PHS (registered trademark), a child does not continuously receive radio waves. Also, for example, even when a plurality of children enter a tunnel (for a game) and do not know who came out first, it is possible to correctly grasp the position information and identify the individual. Further, for example, if the movement amount data is taken out to a server or the like at the entrance of the door, the position of the child until the child passes through the door can be traced from at least the movement amount up to that point and the position measurement result by the TV camera. .
[0021]
Also, the moving object position detection system of the present invention uses an acceleration, speed, angular acceleration, or angular velocity sensor as the movement amount detection means, and uses an image processing camera or a weight sensor as the position detection means. It is characterized by the following.
Thus, for example, a sensor capable of detecting a fine motion, such as an acceleration gyro, can be attached to an infant to obtain data that can be used for a fine evaluation such as the effect of infant education. Further, in position detection by PHS (registered trademark), the more the accuracy of the position is to be improved, the shorter the interval at which radio waves are transmitted, and the more the infant is exposed to radio waves. However, in the present invention, since the position is detected by the TV camera or the like, extremely fine measurement of about 1/30 second can be performed from the beginning.
[0022]
Further, the method for detecting the position of a moving object according to the present invention is a method for detecting a position of a moving object for detecting a continuous position of one or more moving objects. A moving amount detection procedure for continuously detecting the moving amount of the moving object, and the position of each moving object can be individually detected continuously or discontinuously without identifying the moving object itself. The position detection procedure to be performed, the movement amount information detected by the movement amount detection procedure, and the position information of the moving body detected by the position detection procedure are compared with each other, and the continuous position of each moving body is specified by the moving body. And a collation procedure for detection.
Thereby, the positions of a plurality of children can be individually and continuously detected. Further, unlike the case of position detection using PHS (registered trademark), a child does not continuously receive radio waves. Also, for example, even when a plurality of children enter a tunnel (for a game) and do not know who came out first, it is possible to correctly grasp the position information and identify the individual. Further, for example, if the movement amount data is taken out to a server or the like at the entrance of the door, the position of the child until the child passes through the door can be traced from at least the movement amount up to that point and the position measurement result by the TV camera. .
[0023]
Further, the method for detecting the position of a moving object according to the present invention is a method for detecting a position of a moving object for detecting a continuous position of one or more moving objects. A moving amount detection procedure for continuously detecting the moving amount of the moving object, and the position of each moving object can be individually detected continuously or discontinuously without identifying the moving object itself. The position detection procedure to be performed, the movement amount information detected by the movement amount detection procedure, and the position information of the moving body detected by the position detection procedure are collated, and a continuous position of each of the moving bodies is specified. The position detection in the time zone in which the position detection procedure cannot continuously detect the position of the moving object can be detected in the time zones before and after the movement amount detection procedure and the position detection procedure. Based on the collation relationship Only the output movement amount information of the moving amount detection procedure by, characterized in that it comprises a verification procedure for detecting the position of the moving body.
Thereby, the positions of a plurality of children can be individually and continuously detected. Further, unlike the case of position detection using PHS (registered trademark), a child does not continuously receive radio waves. Also, for example, even when a plurality of children enter a tunnel (for a game) and do not know who came out first, it is possible to correctly grasp the position information and identify the individual. Further, for example, if the movement amount data is taken out to a server or the like at the entrance of the door, the position of the child until the child passes through the door can be traced from at least the movement amount up to that point and the position measurement result by the TV camera. .
[0024]
Further, the computer program according to the present invention is provided in a computer in a position detection system of a moving object for detecting a continuous position of one or more moving objects, and a moving amount of the moving object attached to each moving object. Moving amount detection procedure for continuously detecting the position, and position detection that enables the position of each of the moving objects to be individually detected continuously or discontinuously without identifying the moving object itself. And comparing the moving amount information detected by the moving amount detecting step with the position information of the moving object detected by the position detecting step, and detecting successive positions of the moving objects while identifying the moving object. And a program for executing a collation procedure.
[0025]
Further, the computer program according to the present invention is provided in a computer in a position detection system of a moving object for detecting a continuous position of one or more moving objects, and a moving amount of the moving object attached to each moving object. Moving amount detection procedure for continuously detecting the position, and position detection that enables the position of each of the moving objects to be individually detected continuously or discontinuously without identifying the moving object itself. And comparing the moving amount information detected by the moving amount detecting step with the position information of the moving object detected by the position detecting step, and detecting a continuous position of each moving object while identifying each moving object. In addition, the detection of the position in the time zone in which the position detection procedure cannot continuously detect the position of the moving object is performed by comparing the movement amount detection procedure with the position detection procedure in the preceding and following time zones. Based on engagement, a program for executing a verification procedure for detecting the position of the moving object only by the output movement amount information of the displacement detection procedure.
[0026]
The respective means used in the moving object position detecting system according to the first and second aspects operate as follows.
[0027]
The moving amount detecting means includes, for example, a sensor for detecting acceleration and each acceleration, an amplifier for detecting this signal, a band-pass filter, and the like. This sensor data is recorded at a relatively short sampling cycle in a memory means provided inside the movement amount detecting means. However, if there is a non-invasive data transfer means, the data may be transferred from the mobile unit to the server or the like at regular time intervals. If possible, the data may be transferred to a server or the like in real time. If the number of transfers is small, for example, even if wireless communication is used, it is considered that the influence on children is small. Note that the movement amount data is calculated by integrating the output of the acceleration sensor or the like. This calculation may be performed by the movement amount detection means or may be performed by the collation means described later.
[0028]
The position detecting unit is a unit that detects the position of the moving object based on, for example, an image captured by a camera or the like. For example, from a camera image, a fixed background taken in advance is deleted, only a human image is cut out, an individual part is specified as a human, and a plurality of human images are continuously combined. The part determines the position of the human from its form. As will be described later, in the present invention, the position identification of the individual is allowed even if there is no recognition probability of 100%. Note that the detection does not necessarily need to be performed using a camera, and a pressure sensor, an infrared detection device, or the like embedded in a wide area on the floor may be used. It should be noted that it is not possible to recognize who a person is with such position detecting means alone. Further, even if a large motion can be detected, a fine motion cannot be detected.
[0029]
The collation means includes the position detection means (the position of the human is known but the individual cannot be identified) and the movement amount detection means (the individual has been identified in advance and the fine movement can be identified, but due to the influence of the drift, The position of a specific individual is not known at all) and the position is detected while identifying the individual name.
[0030]
Basically, the operation of the above-mentioned matching means is based on a rough temporal change of the moving amount for each individual output from the position detecting means and a temporal change of the moving amount calculated from the output of the moving amount detecting means. In comparison, the two persons are searched for similar ones, and for each person detected by the position detecting means, the individual person is associated with the movement amount detecting means. As a result, the position is identified while identifying the individual name. Is detected.
[0031]
Each means used in the moving object position detecting system according to the third and fourth aspects operates as follows.
[0032]
The moving amount detecting means can be realized with the same configuration as the moving amount detecting means according to the first aspect. For example, it comprises a sensor for detecting acceleration or angular acceleration, an amplifier for detecting this signal, a band-pass filter, and the like. In the embodiment of the present invention to be described later, this record is stored in a memory unit provided inside the movement amount detecting unit at a relatively short sampling cycle. However, if there is a non-invasive data transfer means, the data may be transferred from the mobile unit to the server or the like at regular time intervals. If possible, the data may be transferred to a server or the like in real time. If the number of transfers is small, for example, even if wireless communication is used, it is considered that the influence on children is small. Note that the movement amount data is calculated by integrating the output of the acceleration sensor or the like. This calculation may be performed by the movement amount detection means or may be performed by the collation means described later.
[0033]
The position detecting means can be realized by the same configuration as the position detecting means of the first aspect. For example, it is a unit that detects the position of the moving object based on an image captured by a camera or the like. For example, from a camera image, a fixed background taken in advance is deleted, only a human image is cut out, an individual part is specified as a human, and a plurality of human images are continuously combined. The part determines the position of the human from its form. As will be described later, in the present invention, the location of an individual can be tolerated without a 100% recognition probability. It should be noted that it is not possible to recognize who a person is with such position detecting means alone. Further, even if a large motion can be detected, a fine motion cannot be detected.
[0034]
The collation means includes the position detection means (the position of the human is known but the individual cannot be identified) and the movement amount detection means (the individual has been identified in advance and the fine movement can be identified, but due to the influence of the drift, The position of a specific individual is not known), and the position is detected while identifying the individual.
[0035]
Basically, the operation of the above-mentioned matching means is based on a rough temporal change of the moving amount for each individual output from the position detecting means and a temporal change of the moving amount calculated from the output of the moving amount detecting means. In comparison, the two persons are searched for similar ones, and for each person detected by the position detecting unit, the individual is associated with the moving amount detecting unit. As a result, while the individual name is specified, Position to be detected can be detected.
[0036]
However, the position detecting means cannot always correctly and continuously track the position of one individual. For example, consider two moving objects (individuals) A and B. A and B do not recognize who is a specific individual as a position detecting means. However, it merely detects that there is a human in the place and labels A and B. The problem is whether or not the position detecting means can track A if it is completely correct from the opening time to the closing time. In practice, this may be possible to some extent if all mobile people are always acting at a certain distance. However, this is an unrealistic setting, and actually causes the following problem.
[0037]
"Redundant": When tracking, due to some kind of reflected light, it looks as if there were originally no human beings.
"Disappearance": Despite the fact that there is a human being, human beings cannot be detected temporarily for some reason.
“Tracking mistake”: This is a case where the humans A have contacted / continuous with each other, so that it is not completely known to which of the humans A, which had separated before and after that, moved.
[0038]
In order to cope with such a case, in the present invention, the position information reported by the position detecting means is assumed to be discontinuous (non-continuous). Obviously, if the above "disappearance" occurs, it will be. Further, when "redundancy" occurs, as a result, the recording is lost before and after the portion where the fake individual is detected, which is equivalent to "loss". In addition, if two moving objects merge and separate again, the tracking must be terminated once at the time of separation, and position tracking after separation must be treated like detection of another person. . This is because it may not be possible to determine which has shifted to which only with the image. Therefore, in this case as well, the position information after separation can be handled as if there is "disappearance" at the time of separation.
[0039]
As described above, the position information of the moving object by the image recognition is not only not able to identify the individual, but also it is not possible to track the individual continuously from the beginning to the next again within the measurement time. Of course, there may be cases where such tracking can be realized by chance. However, in many cases, when humans hug each other or dance while holding hands (persons cannot be identified) The tracking record, which until then was thought to have been tracking a single infant, was broken. Also, for example, the same phenomenon occurs when the user goes through a tunnel in a garden or happens to have an obstacle between a camera and a subject. If two children enter the tunnel, the child who exits the tunnel first cannot be identified as the child who entered the tunnel.
[0040]
The position detecting means has a function of detecting individual moving objects, but causes the above-mentioned "redundancy", "loss", and "tracking error" detection errors. In order to avoid this, as a position detection function, "generation" for detecting a new moving object, "disappearance" for detecting disappearance of a moving object, and "change" for detecting a change from a plurality of divided moving objects to one moving object. Four types of “combination” and “division” for detecting separation of one moving object into a plurality of moving objects are provided, and a segment from “generation” or “division” to “disappearance” or “combination” is defined as a segment. Then, for each segment, the individual moving objects are managed individually (although the moving objects cannot be specified).
Such a case clearly requires two functions. The first matching function is a task of identifying each individual segment by comparing each piece of position information of the moving object individually detected for each segment with the movement amount information. And this gives us the location information from the beginning to the end of a particular individual, but unfortunately it is short in time, but some of it may be missing (for example, And time in the tunnel). The second matching function is a function of checking the position of such a missing time zone. In such a case, a simple solution is to simply perform linear interpolation. Of course, secondary and tertiary multidimensional function interpolation may be performed. That is also included in the gist of the present invention. However, a more accurate means is to fill in this missing portion in a short time by using the position change information obtained as a result of the movement amount detecting means. This is because, within the range of a short-time motion, even a piece of acceleration information from which a DC component has been removed as a measure against drifting has a certain meaning.
[0041]
With the above configuration, even if the output of the position detecting means is not completely continuous, the position of the tracked moving object (actually, in the example of this specification, a human being and an infant) is determined at the time of opening the park. Can be specified continuously from to the time of closing. That is, it is possible to realize personal identification that cannot be obtained only from image information of a camera or the like, and to obtain both detailed information of an acceleration sensor and the like and position information.
[0042]
BEST MODE FOR CARRYING OUT THE INVENTION
[First Embodiment]
FIG. 1 shows a system configuration example of a position detection system 1 according to a first embodiment of the present invention. A moving object position detection system 1 according to the present invention includes a moving amount detecting means 10 attached to the moving object 100 to detect the acceleration, speed, angular acceleration, or angular velocity of the moving object 100, and a moving image detecting means 10 for detecting an image captured by a TV camera or the like. Position detecting means 11 for detecting the position of the moving body 100 without specifying the name or the like of the moving body 100 based on the moving amount information detected by the moving amount detecting means 10 and the position detecting means 11 Each time the position information of the moving object is collated with the position information of the moving object 100, the moving object 100 is identified by the collating means 12 which detects the continuous position while identifying the moving object 100. The moving body 100 is, for example, a human being such as an infant or a nursery teacher or a kindergarten teacher. However, this may be living or inanimate. In particular, it is not limited to being a human.
[0043]
(1) Moving amount detecting means
The movement amount detecting means 10 is mounted on each infant, and is a means for detecting the movement of the infant using an acceleration, speed, angular acceleration, angular velocity sensor or the like. FIG. 3 is a diagram illustrating a configuration example of the movement amount detection unit 10. As shown in FIG. 3, a three-dimensional acceleration sensor 200, an amplifier 201 for amplifying the output of the three-dimensional acceleration sensor 200, a sampling circuit 202 for sampling the output of the three-dimensional acceleration sensor 200 at regular time intervals, and a sampling circuit It comprises a memory circuit 203 for storing the result.
The acceleration sensor 200, the amplifier 201, and the sampling circuit 202 are each selectively used based on a design judgment of a normal engineer. Similarly, the memory circuit 203 includes a nonvolatile memory such as a CMOS-RAM, a flash RAM, an FRAM (ferroelectric memory), and an MRAM (magnetic memory), and a read / write control circuit therefor. Also, an interface circuit 204 for taking out the recorded information to the outside is required. For example, at the end of a day, a nursery teacher or kindergarten teacher removes the device containing the acceleration sensor from the children and attaches it to the device for data transfer. As the interface circuit 204, for example, an RFID (Radio Frequency IDentification) can be used.
Note that the acceleration sensor is not necessarily a three-dimensional sensor. In the case of a three-dimensional or two-dimensional acceleration sensor, a circuit for calculating the absolute value of the sensor output may be added thereto. If necessary, the speed / movement amount calculation circuit 205 may be used to integrate the acceleration to obtain the speed / movement amount data and use it.
[0044]
In the present embodiment, a configuration example in which the signal of the acceleration sensor 200 is transmitted from an infant in real time is not illustrated. However, a transmission circuit can be further added to the circuit configuration shown in FIG. In addition, instead of transmitting data at all times, transfer data to an external server or the like at regular time intervals or at the moment of passing the entrance of the room, for example, by wireless, infrared, ultrasonic, etc. links. You can also.
[0045]
Further, for example, on-demand data transfer in which data is transmitted only when a parent requests a position detection of an infant is conceivable. In this case, since the data is transferred after the parent has agreed, no trouble occurs. If the guardian does not want to detect the position, no radio wave is generated.
[0046]
(2) Position detecting means
The position detecting means 11 is a means for "detecting the approximate position of an individual without specifying the infant (infant name) to be detected", and does not necessarily require 100% recognition performance. As a specific configuration example, for example, the position of the infant is extracted from an image taken from the upper part of a room or a garden by a TV camera.
[0047]
FIG. 4 is a diagram illustrating a configuration example of the position detection unit 11. The position detecting means 11 calculates a difference between the TV camera 300, the image of the TV camera 300 and a background image 301 when no infant is imaged in advance, and a difference circuit 302 for extracting an area occupied by the infant. And a cutout circuit 303 for cutting out each individual region. The final center position of each infant is output from the output circuit 304 as XY coordinates.
[0048]
FIG. 5 is a diagram schematically showing a processing flow of the position detecting means 11 (see FIG. 4). Although the captured image 400 of the TV camera 300 is a normal TV image, a difference extraction result 401 is obtained by a difference from the background image 301. Then, as the image 402 obtained by cutting out the infant area, the center (center of gravity) of the infant area on the image is detected and used as coordinate data.
[0049]
Here, the image processing can be executed at very short time intervals. Since the TV camera shoots 30 times per second even in non-interlaced conversion, the above-mentioned infant region can be extracted every 1/30 second. Thus, one infant area does not suddenly go to another location to warp just because the screen has changed. Therefore, in fact, by notifying the overlap of each extracted infant area, the center data of the infant area is labeled, and the coordinate data belonging to the same label is output as temporally continuous data. You can do it.
[0050]
However, such a continuous labeling is not always possible. When the infant enters the tunnel (for games) and comes out of another place, naturally, the observed data becomes different data. In this case, there is no other choice but to label the image of the infant coming out of the tunnel. Because they may have been replaced by another child in the tunnel.
[0051]
A similar discontinuity in labeling occurs, for example, when two children are hugging each other, wrestling with each other, or overlapping napping naps. For example, when two children embrace each other, it is difficult to identify which part of the two labels belong to the unified child area, even if it is known. Further, when the union regions are separated and moved as individual individuals, the specific labels cannot be given just because the regions overlap. Need to give another label.
[0052]
As described above, basically, the extracted position coordinates are temporally continuous data for each label. However, it does not continue consistently from the start to the end of operation of the system. It is only continuous data from a certain start time to an end time. As described above, a series of position coordinate data to which a certain label is given is referred to as a segment in this specification.
[0053]
FIG. 6 schematically shows the segmented position coordinate data as one-dimensional. Here, six segments are detected. The segments # 1 and # 2 that existed from the beginning disappear in a manner that # 2 merges, and then are divided again into two new segments # 3 and # 4. It continues to exist, and # 6 indicates that it is generated from the middle and disappears from space again so as to disappear. FIG. 6 is a schematic diagram, and the actual position coordinate data is two-dimensional or three-dimensional. Also, as will be confirmed again here, this label was given by the system without permission, and it is not known at all who the infant is. In addition, even if the segment data is located in a temporally distant place and is given a different label, in many cases, the same infant may actually be targeted. At the stage, they cannot be identified.
[0054]
(3) Matching means
FIG. 7 is a diagram for explaining a basic function of the matching unit 12. Here, for the sake of simplicity, it is assumed that there are three above-mentioned segments and that there are three outputs of the movement amount detecting means 10. Although the segments start and end at the same time in FIG. 7, it is not necessary to be in such a segment state. The basic function of the matching means 12 is to compare the segment data with the movement amount data and associate the closest data based on some distance. In FIG. 7, in the movement amount data, the distance obtained by integrating the output of the three-dimensional accelerometer is displayed as the absolute value of the three-dimensional vector (the magnitude of the vector). However, in the present invention, the absolute value is always used. It is not necessary to perform acceleration, and it is not an essential requirement to integrate acceleration into velocity.
Note that a segment does not always have a corresponding moving object. For example, suppose that a certain segment actually detected a cat. Even in this case, the segment having no corresponding movement amount information can be removed as noise because the corresponding moving body is not determined by the operation of the matching unit 12.
[0055]
FIG. 8 is a diagram illustrating a specific configuration example of the matching unit 12. Here, the time-series segment data and three time-series movement amount data are input to the matching unit 12. The segment data is first input to the moving distance calculation circuit 500. Here, the amount of movement of the infant is vectorized by calculating the difference between each element to determine how much movement has occurred before and after the sampling point. Since the original segment data is the position information, the difference was calculated to make it the amount of movement. As a result, segment data corresponding to n sampling points is obtained. The corrected three segment data and the three time-series movement amount data are transmitted to the distance calculation circuit 501. However, it is assumed that there are n time sampling points at the same timing. In other words, both the segment data (after modification) and the movement amount data are vectors of numerical elements having the same n-dimensional number of dimensions.
[0056]
The segment data (after modification) and the movement amount data are input to the distance calculation circuit 501. The distance calculation circuit 501 calculates, for example, the cosine of the angle formed by the segment data and the vector of the movement amount data. In short, it is a correlation coefficient. As a result, a correlation coefficient closer to 1 can be obtained for a pattern with a similar movement amount. The data whose segment data and the moving amount data are most similar have a large correlation coefficient.
[0057]
The calculation result of the correlation coefficient is sent to the association circuit 502. Here, for example, the segment data is associated with the movement amount data by the following procedure.
(A) Step 1: The segment data having the largest positive correlation coefficient and the movement amount data are extracted and associated. The segment data for which the association has been completed and the correlation coefficient data related to the movement amount data are ignored in subsequent steps.
(B) Step 2: The above step 1 is repeated until there is no more segment data and movement amount data.
[0058]
By the above processing, the segment data most similar in pattern and the movement amount data are associated with each other. The movement amount data is acquired for each individual infant. That is, the movement amount data specifies an individual. On the other hand, although the individual cannot be specified in the segment data, the individual name of the segment data is determined by this collation.
[0059]
With the configuration described above, according to the present invention, the position of a child can be continuously acquired without using an invasive signal transmission unit such as a radio wave. Further, according to the present invention, since detailed information such as acceleration can be obtained, fine motion data of a child can be obtained. Further, according to the present invention, since it is not necessary to strike a large facility unlike PHS (registered trademark), position detection can be performed more economically.
[0060]
Further, even when the position information is required to some extent in real time, it is not necessary to report the position by emitting radio waves frequently as in PHS (registered trademark). If data is transmitted at a time when a point is reported at a long time interval or when passing through a door, it is possible to suppress the use of invasive media such as radio waves and reduce the risk of exposure to radiation. it can. Further, when such data transfer is used, for example, the data link means can be easily configured as compared with a method using an RFID tag, and the system price can be reduced.
[0061]
[Second embodiment]
FIG. 2 shows a system configuration example of a position detection system 2 according to a second embodiment of the present invention. A moving body position detection system 2 is attached to each of two or more moving bodies 100 and detects a moving amount detecting means for detecting acceleration, velocity, angular acceleration, or angular velocity of the plurality of moving bodies 100, respectively. 10, position detecting means 11 for detecting the position of the moving body 100 based on a photographed image of a camera or the like without specifying the moving body, moving amount information detected by the moving amount detecting means 10, The position information of the moving object is compared with the position information of the moving object detected by the detecting means 11, and the position of the moving object is erroneously recognized or the position of the moving object is not detected. And a matching means 20 for detecting.
[0062]
The moving body 100 is the same as the moving body 100 shown in FIG. For example, a person such as an infant or a nursery teacher or a kindergarten teacher. However, it can be living or inanimate, and in particular need not be human.
[0063]
(1) Moving amount detecting means
The moving amount detecting means 10 is the same as the moving amount detecting means 10 shown in FIG. It is a means for detecting the movement of the infant, such as an acceleration, speed, angular acceleration, or angular velocity sensor, to determine the amount of movement, which is attached to each infant.
[0064]
(2) Position detecting means
The position detecting means 11 is the same as the position detecting means 11 shown in FIG. "This is a means for detecting the position without specifying the name of the infant to be detected. As a specific configuration example, for example, an image taken from the upper part of a room or a garden by a TV camera is used to detect the position of the infant. This is a means for extracting the position.
[0065]
(3) Matching means
The basic function of the matching means 20 includes the function of the matching means 11 shown in FIG. However, the collating unit 20 can perform processing even when the segment data includes extra segment data, partial loss of segment data, separation of merged segments, and the like, as shown in FIG. In FIG. 9, the numbers with “#” are the segmentation labels. However, in the case of separation of the merged segments, as a result, although it is certain that the segment is continuous with the segment before the merge, another segment label is assigned because it is unknown which one is.
[0066]
The following steps are executed as the processing of the matching unit 20.
(A) Step 1: For example, in accordance with the steps described for the collating means 12, each segment is associated with the movement data that has been individually identified. That is, personal identification information is assigned to each segment. In FIG. 9, "Taro-kun" and "Hanako-san" are the personal identification results.
[0067]
(B) Step 2: The segment data belonging to the same individual is made continuous according to the individual identification result. At this time, in the case where there is a leap in time, interpolation is performed based on the movement amount calculated from the movement amount data. This is because, if the detected data is three-dimensional acceleration data, the speed is obtained by integrating the three-dimensional acceleration data, so that the movement amount can be calculated by multiplying the speed and the time.
Also, in a state where two or more segments are merging (that is, a state where two or more children are hugging and cannot be distinguished from each other, but where the group is known), the position coordinates during the merging are calculated. And is diverted as the positional information of the children belonging to both segments identified as the result.
[0068]
(C) Step 3: Repeat steps 1 and 2 above for all children until the segments are continued. In these processes, a segment that does not belong to anywhere and has a relatively short duration is a false segment that appears due to erroneous recognition and is excluded from personal identification processing.
[0069]
Through the above processing, for example, even when a child enters a tunnel and does not know which one comes first, the position information can be correctly grasped and personal identification can be performed. In addition, even in places where the position of the TV camera is not known, such as in a tunnel, or in cases where shooting by the TV camera is skipped, such as in a toilet, the position information that is not obtained can be moved. It can be supplemented from the amount detection result.
[0070]
Since the present invention intentionally avoids using an invasive transmission medium such as radio waves, the example described above does not consider frequent transmission of movement amount data. On the other hand, the position detection by the TV camera is executed in real time. For example, if data can be taken out to a server or the like at the entrance of the door, the position of the child up to passing through the door can be traced from at least the movement amount up to that point and the position measurement result by the TV camera.
[0071]
Note that this is possible with the present invention as a result of the child's personal identification at the moment of entering the door. For this reason, as long as the position measurement result by the TV camera can be measured in a single segment thereafter, the position of the target child (the telemetry of the movement amount data has not been completed yet). Nevertheless). For example, even when a parent wants to observe his / her child as an image from a remote place using a Web camera or the like, it is necessary to take a picture of only his / her child who is surely looking for the parent in a large number of rooms or a large room. Can be transmitted.
[0072]
Of course, as described above, it is technically possible to transmit a TV image with respect to the specified child's position if real-time position detection using PHS (registered trademark) is performed. However, for that purpose, it is necessary to output a radio wave from the PHS (registered trademark) in a short cycle to realize processing (by exposing a child to the radio wave). In the present invention, on the other hand, even if telemetry by radio waves is performed, the frequency is low, and facilities such as base stations and RFIDs are extremely simple.
[0073]
The embodiment of the present invention has been described above. However, each function in the movement amount detecting means 10, the position detecting means 11, the matching means 12 and 20 shown in FIGS. 1 and 2 is realized by dedicated hardware. And, if possible, a general-purpose information processing device such as a memory and a CPU (Central Processing Unit), and load a program (not shown) for realizing the function of this processing unit into the memory. Then, the function may be realized by executing the program.
[0074]
Further, a program for realizing the functions in the movement amount detecting means 10, the position detecting means 11, the collating means 12 and 20 shown in FIGS. 1 and 2 is recorded on a computer-readable recording medium, and is recorded on the recording medium. Necessary processing may be performed by causing the computer system to read and execute the recorded program. Here, the “computer system” includes an OS and hardware such as peripheral devices.
[0075]
The “computer-readable recording medium” refers to a general medium such as a flexible disk, a magneto-optical disk, a ROM, a CD-ROM, and a storage device such as a hard disk built in a computer system.
Further, a “computer-readable recording medium” refers to a communication line for transmitting a program via a network such as the Internet or a communication line such as a telephone line, which dynamically holds the program for a short time. In this case, a medium that stores a program for a certain period of time, such as a volatile memory in a computer system serving as a server or a client in that case, is also included.
Further, the above-mentioned program may be for realizing a part of the above-mentioned functions, and may be for realizing the above-mentioned functions in combination with a program already recorded in a computer system, that is, a so-called difference file (differential Program).
[0076]
As described above, the embodiments of the present invention have been described. However, the moving object position detecting system of the present invention is not limited to only the above-described illustrated examples, and various changes may be made without departing from the gist of the present invention. Of course, it can be added.
[0077]
It should be noted that the collation of the movement amount information obtained from the movement amount detection means with the position information obtained from the position detection means does not necessarily have to be performed by each means independently of the other means. . For example, when position extraction from a camera image is used as the position detection means, one realization method is to take the difference in a pixel unit of a continuously shot image to determine that something is moving. This is a technique for extracting the occupied area.
Needless to say, the position information obtained by the difference detection is not a point but an area. The portion corresponding to a certain moving object includes, for example, a shadow portion, and is not always equal to the appearance of the moving object itself (in other words, the outline of the appearance seen from the camera). In addition, differences are detected in areas that have been shaded up to now because difference signals from multiple moving objects overlap, or curtains sway in the wind, or sun rays suddenly enter. It may be extracted as a region that has been changed.
[0078]
On the other hand, the movement amount detecting means detects the movement amount of each individual after the individual is specified. For example, when an acceleration sensor is used as the movement amount detecting means, the technical limit of the acceleration sensor (the frequency range in which acceleration can be detected is limited, or the DC component is superimposed due to the DC drift of the sensor amplifier) Therefore, even if the acceleration is integrated due to the fact that the DC output component cannot be directly treated as the acceleration, the actual speed will not be accurate. However, the performance of acceleration sensors continues to improve with the development of micromachine technology in recent years, and even in the current situation, in a very short time, by integrating the information of the acceleration sensor, the moving speed and moving distance can be increased to some extent. Can be estimated.
[0079]
For example, when the moving object is not moving at all, there is no signal from the acceleration sensor, but if the lighting conditions of the camera change during that period (during the period when the moving object is not moving), a difference signal appears on the entire screen. I do. However, since there is no movement signal from the moving body, this need not be detected as a difference signal output in the first place. As described above, if the information from the movement amount detecting means is used together, the extra calculation can be eliminated.
[0080]
In addition, it is assumed that only a specific moving body has moved while the lighting conditions are changing on the screen. In this case, the difference signal appears on the entire screen. Although it includes a portion resulting from the movement of the specific moving object, most is a portion caused by a difference in lighting conditions. What is really needed is only the difference information of the part where the moving body is moving. In such a case, it is necessary to extract only a partial area that moves in accordance with the movement information from the image information of the camera from the movement information of the moving body to obtain a difference signal. Otherwise, if the difference is obtained only from the image data of the camera, both the part where the moving body is moving and the part that has become the difference due to the illumination change become the position information. If such position information whose information is too redundant is used, the collation itself becomes difficult.
[0081]
As described in the above example, the gist of the present invention includes utilizing information of the movement amount detecting means in calculating the position of the position detecting means. It is rather natural to use another amount of information from the stage of segmenting the area to improve the performance of the position detecting means on the pattern recognition technology, and the present invention includes the processing.
[0082]
【The invention's effect】
As described above, in the moving object position detecting system and moving object position detecting method of the present invention, the moving amount of each moving object is detected by the moving amount detecting means such as an acceleration sensor, and the position of each moving object is detected. Is continuously or discontinuously detected individually without identifying the moving object itself by position detection means such as a TV camera, and the movement amount information obtained by the movement amount detection means and the position obtained by the position detection means. By collating with the information, the continuous position of the moving body is detected while identifying the moving body.
Thereby, the positions of a plurality of children can be individually and continuously detected. Further, unlike the case of position detection using PHS (registered trademark), a child does not continuously receive radio waves. Also, for example, even when a plurality of children enter a tunnel (for a game) and do not know who came out first, it is possible to correctly grasp the position information and identify the individual. Further, for example, if the movement amount data is taken out to a server or the like at the entrance of the door, the position of the child until the child passes through the door can be traced from at least the movement amount up to that point and the position measurement result by the TV camera. .
[0083]
Further, in the position detection system for a moving object of the present invention, the movement amount detection means uses an acceleration, speed, angular acceleration, or angular velocity sensor, and the position detection means uses an image processing TV camera or a weight sensor. use.
Thus, for example, a sensor capable of detecting a fine motion, such as an acceleration gyro, can be attached to an infant to obtain data that can be used for a fine evaluation such as the effect of infant education. Further, in position detection by PHS (registered trademark), the more the accuracy of the position is to be improved, the shorter the interval at which radio waves are transmitted, and the more the infant is exposed to radio waves. However, in the present invention, since the position is detected by the TV camera or the like, extremely fine measurement of about 1/30 second can be performed from the beginning.
[0084]
In the moving object position detecting system and the moving object position detecting method according to the present invention, the moving amount of each moving object is detected by a moving amount detecting means such as an acceleration sensor, and the position of each moving object is detected by a TV camera or the like. Even if the moving body itself is not identified by the position detecting means, the moving body itself is continuously or discontinuously detected individually, and the moving amount information detected by the moving amount detecting means and the position information of the moving body detected by the position detecting means And detecting the continuous position of the moving body while identifying the moving body, and detecting the position in the time zone in which the position detecting means cannot continuously detect the position of the moving body includes the moving amount detecting means. Then, based on the collation relationship detected by the position detecting means in the preceding and following time zones, the position of the moving object is detected only by the moving amount information output from the moving amount detecting means.
Thereby, the positions of a plurality of children can be individually and continuously detected. Further, unlike the case of position detection using PHS (registered trademark), a child does not continuously receive radio waves. Also, for example, even when a plurality of children enter a tunnel (for a game) and do not know who came out first, it is possible to correctly grasp the position information and identify the individual. Further, for example, if the movement amount data can be taken out to a server or the like at the entrance of the door, the position of the child until the child passes through the door can be traced from at least the movement amount up to that point and the position measurement result by the TV camera.
[Brief description of the drawings]
FIG. 1 is a diagram illustrating a system configuration example of a position detection system according to a first embodiment of the present invention.
FIG. 2 is a diagram illustrating a system configuration example of a position detection system according to a second embodiment of the present invention.
FIG. 3 is a diagram illustrating a configuration example of a movement amount detection unit.
FIG. 4 is a diagram illustrating a configuration example of a position detection unit.
FIG. 5 is a diagram schematically showing a flow of processing in a position detecting means.
FIG. 6 is a schematic diagram of segmented position coordinate data.
FIG. 7 is a diagram for explaining a basic function of a matching unit.
FIG. 8 is a diagram illustrating a configuration example of a matching unit;
FIG. 9 is a schematic diagram of segment data.
[Explanation of symbols]
1,2 Moving object position detection system
10 Moving amount detecting means
11 Position detecting means
12,20 collation means
100 mobile
200 acceleration sensor
201 amplifier
202 Sampling circuit
203 memory circuit
204 interface circuit
205 Speed / movement data calculation circuit
300 TV camera
301 Background Image
302 Difference circuit
303 Cutout circuit
304 output circuit
400 photographed images
401 Difference extraction result
402 Cutout image
500 Travel distance calculation circuit
501 Distance calculation circuit
502 Mapping Circuit

Claims (8)

A position detection system of a moving body for detecting a continuous position of one or more moving bodies,
A moving amount detecting means attached to each of the moving bodies to continuously detect a moving amount of the moving body,
The position of each of the moving bodies, without identifying the moving body itself, a position detecting means that can be individually detected continuously or discontinuously,
Collation for comparing the movement amount information detected by the movement amount detection means with the position information of the moving body detected by the position detection means, and detecting continuous positions of the moving bodies while identifying the moving body. Means for detecting the position of a moving object. As the movement amount detection means, acceleration, speed, angular acceleration, or using an angular velocity sensor,
The position detection system according to claim 1, wherein an image processing camera or a weight sensor is used as the position detection unit. A position detection system of a moving body for detecting a continuous position of one or more moving bodies,
A moving amount detecting means attached to each of the moving bodies to continuously detect a moving amount of the moving body,
The position of each of the moving bodies, without identifying the moving body itself, a position detecting means that can be individually detected continuously or discontinuously,
The moving amount information detected by the moving amount detecting means and the position information of the moving body detected by the position detecting means are compared with each other, and a continuous position of each moving body is detected while identifying each moving body, The detection of the position in the time period in which the position detecting means cannot continuously detect the position of the moving body is performed based on the collation relationship detected by the moving amount detecting means and the position detecting means in the preceding and following time periods. And a collating means for detecting the position of the moving object only based on the moving amount information output from the moving amount detecting means. As the movement amount detection means, acceleration, speed, angular acceleration, or using an angular velocity sensor,
The position detection system according to claim 3, wherein an image processing camera or a weight sensor is used as the position detection unit. A position detection method of a moving body for detecting a continuous position of one or more moving bodies,
A movement amount detection procedure attached to each of the moving bodies to continuously detect the movement amount of the moving body,
A position detection procedure that enables the position of each of the moving objects to be detected continuously or discontinuously individually without identifying the moving object itself,
Collation for comparing the movement amount information detected by the movement amount detection procedure with the position information of the moving body detected by the position detection procedure, and detecting continuous positions of the moving bodies while identifying the moving body. And a procedure for detecting the position of the moving object. A position detection method of a moving body for detecting a continuous position of one or more moving bodies,
A movement amount detection procedure attached to each of the moving bodies to continuously detect the movement amount of the moving body,
A position detection procedure that enables the position of each of the moving objects to be detected continuously or discontinuously individually without identifying the moving object itself,
The moving amount information detected by the moving amount detecting step and the position information of the moving object detected by the position detecting step are compared with each other, and the continuous position of each moving object is detected while identifying each moving object, The detection of the position in the time zone in which the position detection procedure could not continuously detect the position of the moving object is based on the collation relationship in which the movement amount detection procedure and the position detection procedure were able to detect in the preceding and following time zones. And a collation procedure for detecting the position of the moving object only based on the moving amount information output in the moving amount detecting procedure. A computer constituting a position detection system of a moving object for detecting a continuous position of one or more moving objects,
A movement amount detection procedure attached to each of the moving bodies to continuously detect the movement amount of the moving body,
A position detection procedure that enables the position of each of the moving objects to be detected continuously or discontinuously individually without identifying the moving object itself,
Collation for comparing the movement amount information detected by the movement amount detection procedure with the position information of the moving body detected by the position detection procedure, and detecting continuous positions of the moving bodies while identifying the moving body. Steps and programs to execute. A computer constituting a position detection system of a moving object for detecting a continuous position of one or more moving objects,
A movement amount detection procedure attached to each of the moving bodies to continuously detect the movement amount of the moving body,
A position detection procedure that enables the position of each of the moving objects to be detected continuously or discontinuously individually without identifying the moving object itself,
The moving amount information detected by the moving amount detecting step and the position information of the moving object detected by the position detecting step are compared with each other, and the continuous position of each moving object is detected while identifying each moving object, The detection of the position in the time zone in which the position detection procedure could not continuously detect the position of the moving object is based on the collation relationship in which the movement amount detection procedure and the position detection procedure were able to detect in the preceding and following time zones. And a collation procedure for detecting the position of the moving object only based on the displacement information output in the displacement detection procedure.

Images