JP2008304230A - LOCATION SYSTEM, LOCATION METHOD, ITS PROGRAM, AND RECORDING MEDIUM - Google Patents

Abstract

A location system, a location method, a program thereof, and a recording medium capable of satisfactorily estimating the position and direction of an object in combination with a conventional topology estimation system by the inventors of the present invention are provided.
In a location system that estimates the position and direction of an object in the space based on sound information transmitted and received between objects existing in the space, each object emits a sound unique to the object toward the surroundings The device 11 and a recording device 12 for recording an object specific sound emitted from another object sounding device are mounted, and an object specific sound from the sounding device 11 of a certain object is recorded by the recording device 12 of another object. Whether or not it is detected from a plurality of recording device data subjected to time synchronization processing, and detecting a one-to-one distance relationship between objects, a direction relationship, and a correlation consisting of IDs. Location system.
[Selection] Figure 1

Description

  The present invention relates to a location system that estimates the position and direction of an object existing in a space where an information support service is provided.

  In a space with many objects such as humans, robots, and exhibits, it is important to construct an information support system based on position and a man location management system. For example, in an indoor space where many people gather for activities such as parties, social gatherings, and workshops, there is an increasing expectation for realizing these systems.

  On the other hand, in view of the importance of estimating the situation of the interaction information between objects in the space, the inventor of the present invention can obtain this interaction information from the absolute position coordinates indicating the absolute position of each object alone in the space. However, it is useful to use a topology that shows the connection of the whole object in the space obtained from the relationship between the relative position and orientation of the objects, especially if you can estimate not only the positional relationship between the objects but also the orientation relationship, When a user such as a human being or a robot moves in an exhibition or an environmental object, it has been found that various information support more suitable for the user's situation changing from moment to moment can be realized.

  Therefore, there has already been proposed a topology estimation system that can estimate the object topology in the entire space with a simple system configuration even in a space where a large number of objects such as humans and robots exist relatively densely (patent) Reference 1).

  In this topology estimation system, for example, as illustrated in FIG. 11, the environment device 2 estimates the topology of an object in the space based on the object information wirelessly transmitted from the object device 1 of each object existing in the space. The object device 1 includes an ID transmission unit 11 that transmits its own object ID and a transmission direction ID of the object ID to the surroundings, and an object of the other object device 1 transmitted by another surrounding object device 1 An ID receiving unit 12 that receives an ID and a transmission direction ID, and an information transmission unit 13 that transmits object information including its own object ID and the received object ID and transmission direction ID of another object device 1 to the environmental device 2 The environmental device 2 has An information receiving unit 21 that receives object information transmitted by the object device 1, a database unit 22 that stores the received object information, and a relative positional relationship with other objects for each object based on the stored object information, A processing unit 23 for obtaining an object topology from the obtained set data of relative positional relationships is provided.

  According to this topology estimation system, an object device is attached to each object such as a human being, a robot, or an exhibit existing in the target space, and the object device and its transmission direction ID, that is, the object in which direction, between the object devices. Information indicating whether the ID is transmitted (hereinafter, the object ID and the transmission direction ID are collectively referred to as “ID”) is exchanged, and the object information as a result of the exchange is sequentially transmitted from each object device to the environmental device. Based on the collected object information, the relative positional relationship with each other object, that is, the local and relative positional relationship centered on each object is obtained for each object, and space is obtained from the set data of the relative positional relationship. Overall object topology, that is, in space It is possible to estimate the overall and relative positions of all objects to each other.

  More specifically, for example, in a situation where the object devices A, B, C, and D possessed by each object existing in the space where the information support service is provided wirelessly send their IDs to the surroundings. First, when the object device A receives the IDs from the object devices B, C, and D and wirelessly transmits its own object ID and the received other ID to the environmental device, the object device A and the other object ID Based on this, it can be seen that the object device A exists in the surrounding space of the object devices B, C, D, that is, within the range in which the ID transmitted from the object devices B, C, D can be received. The positional relationship with the devices B, C, and D can be determined, and the object device A can detect other objects based on the transmission direction ID. Object device B, C, can tell you receive that originate ID toward any direction from D, thus it can be determined object device B, C, the directional relationship between D. As a result, the relative positional relationship represented by the position and direction between the object devices B, C, and D with the object device A as the center is obtained.

  Similarly, the relative positional relationship with respect to each of the other object devices B, C, D is obtained.

  Then, the relative positional relationships among these object devices A, B, C, and D are aggregated to obtain the relative positional relationship between all the object devices A, B, C, and D in the entire space. This process executes a method for finding the optimal solution that eliminates or reduces the relative position of each relative positional relationship as much as possible, such as recursive operation, EM algorithm, spring model, Bayesian network, genetic algorithm, etc. And you can do it.

  Here, for example, when the object device A has not received the ID from the object device D, information on the relative positional relationship with respect to the object device D around the object device A cannot be obtained. In addition, this shows that the object device A and the object device D are not close to each other when viewed in the entire space.

  Further, for example, the object device A receives the ID from the object device B in front of it, but does not receive the ID from the object device D in front of it, and the object device B is behind the object device B. When both the ID from the device A and the ID from the object device D ahead are received, the relative positional relationship with respect to the object device D with respect to the object device A is unknown, but the object device B is the center. Since the relative positional relationship between the object device A and the object device D is known, the relative positional relationship among the object device A, the object device B, and the object device D can be estimated from these.

  Of course, innumerable cases other than these are conceivable, but as can be understood from the above description, the relative positional relationships of the object devices A, B, C, and D are local relative to each other. The relative positional relationship obtained by collecting and analyzing as described above can be called the overall relative positional relationship between objects in the space, and this can be called the spatial positional relationship. This is an object topology showing the connection of all objects as seen from the whole.

Note that the transmission direction ID included in the ID that the object device wirelessly transmits to the surroundings is information that is preset and stored in the device as to which direction the ID is transmitted from the object device. ID must also travel in that direction in the surrounding space. Therefore, as in the second topology estimation system described above, it is preferable to use an electromagnetic wave that travels straight through the space for ID transmission, so that the ID is accurately transmitted in the direction indicated by the transmission direction ID. The ID is propagated along the direction in the surrounding space. As this electromagnetic wave, one having directivity such as light such as infrared light can be considered. Ultrasound can also be used. For straightness and directivity, for example, when the transmission direction is set to 8 directions around the object device (front / rear left / right, right front / left front, right rear / left rear), There is a need for a degree of advancement that does not interfere with the direction, or even within the range that is acceptable in the topology estimation process even if there is interference.
JP 2006-174185 A

  It is an object of the present invention to provide a location system, a location method, a program thereof, and a recording medium that can satisfactorily estimate the position and direction of an object in combination with the topology estimation system as described above.

  This invention provides the system which can acquire the relationship of the relative distance and relative direction by acoustic analysis as what solves said subject.

More specifically, for example, as illustrated in the functional block diagram of FIG. 1, a location system that estimates the position and direction of an object in the space based on sound information transmitted and received between objects existing in the space. In
Each object is equipped with a sounding device 11 that emits an object-specific sound toward the surroundings, and a recording device 12 that records the object-specific sound emitted from the sounding device 11 of another object.
Whether or not an object specific sound from the sound generator 11 of an object is recorded by the recording device 12 of another object is found from a plurality of recording device data subjected to time synchronization processing, and one-to-one correspondence between the objects. Detecting a correlation consisting of a distance relationship, an orientation relationship, and an ID,
A location system is provided.

Further, in a location method for estimating the position and direction of an object in the space based on sound information transmitted and received between objects existing in the space,
Find out whether or not an object specific sound emitted from a sound device attached to an object is recorded by a recording device attached to another object from a plurality of recording device data that has been time-synchronized. Detecting a one-to-one distance relationship, an orientation relationship, and a correlation between IDs;
A location method is provided.

  The present invention also provides a location program for causing a computer to execute the location method and a computer-readable recording medium on which the program is recorded.

According to the above location system and method of the present invention, the position and direction of an object can be estimated well, and furthermore, it can operate in a congested space What is necessary is just to be able to acquire and the effect that a complicated environmental apparatus is unnecessary is realizable.

  In the present invention, a commercially available IC recorder or IC player can be used as the sound generation device and the recording device, and the above-described effects can be realized without using a special device. Also, since the sound around the object can be recorded, the user can look back on the surrounding environment at that time.

  In addition, real-time position estimation is possible by receiving a large number of sound signals in real time using FM radio waves or the like.

[First embodiment]
FIG. 1 schematically shows an embodiment of the present invention.

  As described above, the position and orientation of the object existing in the target space can be estimated.

  In the embodiment of FIG. 1, first, each object is equipped with an object device 1 including a sounding device 11 and a recording device 12.

  The sound generation device 11 emits an object specific sound toward the surroundings, and includes a sound source unit, a control unit, and the like (not shown) necessary for transmitting the specific sound. Regarding the object unique sound, the object to which the sounding device 11 is attached and the transmission direction from the sounding device 11 are determined in advance, and have the sound and transmission direction unique to the object.

  As the sound unique to the object, for example, a sound having one or both of a unique single or plural frequency characteristics and a unique random noise sound determined in advance for each object, or mixed with these sounds You can consider what you have. For this unique sound, for example, unique music or the like can be adopted for each object, and by transmitting different music respectively assigned from the sound generation device 11, the object can be identified from the characteristic analysis of each music. Further, it is also possible to adopt a unique sound that can be said to be a digital watermark, which is indistinguishable or difficult to be discerned by human ears but can be discriminated that identification information such as ID is included when analyzed by a computer. The watermark sound is generated at the same time as the music, or the digital watermark sound itself is included in the music (in these cases, since the digital watermark sound is unique to the object, the music is the same between objects). However, it is possible to construct a system in which a computer can detect an ID in spite of a normal acoustic environment.

  The recording device 12 records the object specific sound emitted from the sound generation device 11, and includes a recording unit, a control unit, and the like (not shown) necessary for recording the specific sound.

  The sounding device 11 and the recording device 12 may be attached as a single object device 1 or may be separated, but as illustrated in FIG. 2, a plurality of sets are attached to each object. To do. As a result, as illustrated in FIG. 3, it is possible to cover transmission / reception (transmission and recording) of natural sounds in, for example, the four directions before and after the object, with the object as the center. If the number of devices, that is, the direction resolution is high, the accuracy of the estimated position and orientation is improved. It is also preferable to cover all directions as much as possible and reduce the overlap in the direction of the devices. In addition, the direction resolution is improved by shifting the transmission and reception directions.

  Then, a location estimation based on a specific sound transmitted and received between the sounding device 11 and the recording device 12, that is, between the sounding device 11 of a certain object device 1 and the recording device 12 of another object device 1, is performed. Whether or not the object specific sound from the sounding device 11 is recorded by the recording device 12 of another object from the plurality of recording device data subjected to the time synchronization processing, and the one-to-one distance relationship between the objects, This is realized by detecting the correlation between the orientation relationship and the ID.

  Specifically, for example, as shown in the flowchart illustrated in FIG. 4, first, the recording device 12 sequentially records object-specific sounds from the sounding device 11 of other objects located around. This recording can be said to record the activity of the object in the target space, that is, the relationship with other objects (step S1).

  Next, the recorded data is extracted (step S2), and the head alignment time synchronization process between the recorded data is executed (step S3).

  For example, as shown in FIG. 5, when recording data of the user AB as an object is obtained, it is usually difficult to accurately synchronize the recording data. For example, even if the recording buttons of two IC recorders are pressed simultaneously, a deviation of about several milliseconds occurs. Also, at the start of recording, two IC recorders are brought close to each other and a knocking sound is generated at an equal distance from both recorders. Two recording files are recorded from the time when the same sound was recorded by examining the recording of the two recording data. It is also possible to synchronize the times. However, strictly speaking, the crystal oscillator of the IC recorder does not have the same vibration cycle, and a time difference of about several milliseconds occurs after several tens of minutes. At normal temperature, about 30 cm away, it takes 1 millisecond for sonic waves. In the present invention, since it is preferable that the distance accuracy is 30 cm or less, time synchronization of 1 millisecond or less is required.

  Therefore, in the present invention, time synchronization processing is performed as follows.

  First, as shown in FIG. 5, when a unique sound of the sound generation device of user B is transmitted, it is received by the IC recorder of user B and then received by the IC recorder of user A. Similarly, the sound unique to the sound generation device of user B is received by user B after being received by user A. When the time difference between the recording data of the user B and the user A is dt as shown in FIG. 5, the time of the sound that has arrived from the user B to the user A is the sound transmission time + dt, and vice versa. The sound time is the sound transmission time -dt. Therefore, if the time difference of the same sound is examined, 1/2 of the sum is the sound transmission time and 1/2 of the difference is the time difference, so that the two recorded data can be synchronized. Further, the distance between the user A and the user B is obtained from the sound transmission time and the sound transmission speed (determined by the temperature and the atmospheric pressure).

  Subsequently, with respect to the individual recording data after the time synchronization processing, the object ID is detected using the unique sound as a key, and the distance relationship and the orientation relationship between the objects are compared with other recording data to detect (step). S4).

  Here, the ID of the object is detected based on the unique sound, the distance relationship between the objects is detected using the time-of-flight method, and the direction relationship of the object is determined by the microphone directivity of the recording device or the time-of-arrival method. Detected.

  In the object ID detection, the specific sound has characteristics specific to the object. For example, the specific sound characteristics are stored in a database for each object in advance, and the object having the specific sound characteristics positioned with the recording data is searched by database search. What is necessary is just to detect.

  Object distance relation detection is executed by the Time of Flight (ToF) method.

  That is, when the sound transmission time between two objects is found as described above, the distance between the objects can be obtained as follows.

Transmission time × Sound transmission speed The sound transmission speed varies depending on the temperature and the atmospheric pressure, but is approximately 340 m / s, and advances 34 cm in 1 millisecond.

  The object orientation relationship detection is detected using microphone directivity of the recording apparatus or Time of Arrival (ToA) method.

  First, a method using the microphone directivity of the recording apparatus will be described. A microphone as a recording device has directivity by using a directivity type or by being installed on a person (see FIG. 6). Therefore, if the directional characteristics of the microphone corresponding to certain recording data are known, the approximate direction of the user of the specific sound recorded in the data can be determined.

Next, the ToA method will be described as a method using a plurality of microphones. Regarding the orientation relationship, for example, if the sound recorded by the microphones attached to the front and the back of the user A is analyzed, and the specific sound of the user B is detected at the same time, the user B is just left (or right) of the user A. Direction). Explaining precisely when the front and rear microphones are installed as an example, if the previous microphone was consulted dt milliseconds earlier as shown in FIG. 7, the distance between the front and rear microphones is 1, and the right direction from the front Assuming that user B exists in the angle direction of θ, the equation of lcos θ = v · dt (v is the speed of sound) is established. Therefore, the direction of the user B is obtained as θ = cos ⁻¹ (v · dt / l). When the number of microphones is 2, it cannot be determined whether θ = 90 degrees or θ = −90 degrees as described above. However, if three or more microphones are arranged, it can be uniquely obtained. Information of this direction is written in the recording data corresponding to the two microphones.
It is also effective to use various sound source direction estimation methods using a plurality of microphones.

  And after finishing each said process about all the recording data (step S5-Yes), the one-to-one distance relationship and orientation relationship between objects are detected (step S6).

  That is, two pairs of recorded data whose relationship between distance and orientation is found are recorded in data corresponding to each object and orientation. Specifically, for example, the local relationship estimation result that the user B exists at the position about 1 m behind the user A and faces the left side is generated.

  Finally, the topology of the object is estimated by the topology estimation technique described in Patent Document 1 by the inventor of the present invention (step S7).

  Specifically, the global positional relationship can be estimated by creating the above local relationship for all recorded data pairs and then inputting the local relationship as input data to the conventional topology estimation technique. That is, the local positional relationship is obtained in step 6, but when estimating the global positional relationship (position and orientation) of all users and information kiosks and the like, a conventional topology estimation method is used.

  The above processing is executed by the processing device 2 in the embodiment of FIG. The processing device 2 includes an information receiving unit 21, a database unit 22, and a processing unit 23, and recording data is transmitted from the recording device 12 of the object device 1 attached to each object or the separate information transmission unit 13. The information receiving unit 21 receives the data and collects the data in a storage unit such as the database unit 22. Based on the collected data, the processing unit 23 executes each process as described above.

The above processing can be performed by analyzing data recorded by an IC recorder or the like, but the sound received by each microphone is directly transmitted to the server PC by an FM transmitter or the like and processed in real time. It is also possible to estimate the topology with
[Second Embodiment]
By the way, if the position and orientation of the object can be estimated as described above, it is possible to reproduce a sound at an arbitrary intermediate position in the target space as illustrated in FIG. That is, a sound at a designated position is complemented and created from a plurality of recording data recorded at different positions and the recording position by using a sound morphing technique.

  Specifically, where the recording data acquired in the first embodiment is recorded and in which direction the recording data is individually identified and associated with each other from the topology estimation result. The time synchronization between the recorded data at this time is performed using the correlation of the recording time of the specific sound as described above.

  More specifically, the user inputs the time and place that he / she wants to hear. The system selects audio information close to the time and place specified by the user. The audio information is merged so that the ratio increases as the distance to the designated location becomes shorter. This reproduces the approximate sound at the specified location.

  As described above, based on the positional relationship between the recording position and the designated position, the sound that would have been heard there can be complementarily generated using a morphing technique using a plurality of recording data.

FIG. 9 is a conceptual diagram for explaining the process of visualizing the recorded sensor information when the present invention is applied to a workshop that is divided into groups and jointly creates painting works. In the upper left, position information indicating how the user has moved in the venue where desks and information kiosks are arranged is displayed according to the time determined by the time slide bar displayed below. An IC recorder and an IC speaker are set for each user, desk, and information kiosk, and a history of position and orientation is displayed from the recorded audio information of the IC recorder. On the right side, acoustic information corresponding to the time slide bar is reproduced. Since this workshop also included videos and photos, this information is displayed on the upper right screen. The user can change the position of the user according to the time of the past workshop and the conversation, voice, video, and photograph at that time. This is an example of application of the present invention.
[Third embodiment]
Here, the time synchronization processing of the plurality of recording data necessary for realizing the first and second embodiments will be described in detail.

  In order to realize the first and second embodiments, time synchronization between data recorded by a plurality of recording devices 12 is established. This is because, when there are a plurality of wristwatches, the data recorded at the same time by a plurality of recording devices 12 is affected by the shift of the clock used for sampling so that the time progress is slightly different. Are slightly different. Therefore, time correlation detection between the proper sound and the recording data by the ToF method is performed.

  More specifically, for example, as illustrated in FIG. 10, a temporal relationship between recorded data is defined by a combination of detection of a natural sound being transmitted and the ToF method.

  First, when the transmission time of the proper sound A (“proper sound transmission point” in the figure) and the reception time of the proper sound A (“proper sound reception point” in the figure) are close in time, compared to the speed of the sound Since the moving speed of the object is small, it is considered that the conversion of the relative distance is small.

  Next, paying attention to the slope of the line connecting the transmission point and the reception point of the natural sound A, the inclinations of the lines of a certain transmission point and reception point and the next transmission point and reception point are detected, and the inclinations are the same. Is determined to be a position where two data are synchronized in time.

The functional block diagram for demonstrating this invention. The conceptual diagram which shows an example of a sounding device and a recording device. The conceptual diagram for demonstrating the natural sound transmission / reception by a sounding device and a recording device. The flowchart for demonstrating this invention. The conceptual diagram for demonstrating a time synchronous process. The conceptual diagram for demonstrating an object direction relationship detection process. Another conceptual diagram for explaining object orientation relation detection processing. The conceptual diagram for demonstrating another example of this invention. The conceptual diagram for demonstrating an example at the time of applying this invention to a certain workshop. The conceptual diagram for demonstrating another example of this invention. The functional block diagram for demonstrating the conventional system.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Object apparatus 11 Sound generation apparatus 12 Recording apparatus 13 Information transmission part 2 Processing apparatus 21 Information reception part 22 Database part 23 Processing part

Claims (10)

In a location system that estimates the position and direction of an object in the space based on sound information transmitted and received between objects existing in the space,
Each object is equipped with a sounding device that emits object-specific sounds toward the surroundings, and a recording device that records object-specific sounds emitted from sounding devices of other objects,
Whether or not an object specific sound from a sound generator of an object is recorded by a recording device of another object is found from a plurality of recording device data subjected to time synchronization processing, and a one-to-one distance relationship between objects , Detect the correlation consisting of the orientation relationship and ID,
A location system characterized by that. The object ID is detected based on the unique sound.
The location system according to claim 1. The specific sound emitted from the sounding device is determined in advance based on the object to which the sounding device is attached and the transmission direction from the sounding device.
The location system according to claim 2. The natural sound has one or both of a sound having a specific frequency characteristic predetermined for each object and a specific random noise sound, or a sound having a specific frequency characteristic predetermined for each object and a specific random 4. The location system according to claim 3, wherein the location system is mixed with one or both of the noise sounds. The distance relationship between objects is detected using the Time of Flight method.
The location system according to claim 1. The object orientation relationship is detected using the microphone directivity of the recording device or the Time of Arrival method.
The location system according to claim 1. Time-synchronization processing is performed on the recording device data that is recorded by the recording device, and the object specific sound from the sounding device of one object is recorded by the recording device of another object from the recording device data. Further having a processing device to find out whether or not
The location system according to claim 1. In a location method for estimating the position and direction of an object in the space based on sound information transmitted and received between objects existing in the space,
Find out whether or not an object specific sound emitted from a sound device attached to an object is recorded by a recording device attached to another object from a plurality of recording device data that has been time-synchronized. Detecting a one-to-one distance relationship, an orientation relationship, and a correlation between IDs;
A location method characterized by that.   A location program for causing a computer to execute the location method according to claim 8.   A computer-readable recording medium on which the location program according to claim 9 is recorded.