JP2014232421A - Human search system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To effectively use a feature of an on-vehicle camera that a clear, homogeneous face image can be obtained irrespective of a vehicle type.SOLUTION: Driver face image information is generated by picking up a face image of a driver by an on-vehicle camera provided in front of a driver's seat while a vehicle is traveling, and the driver face image information and vehicle position information are transmitted to a database. The database stores therein search target face image information based on a face image of a search target in advance. If the driver's face image information matches the search target face image information, the vehicle position information transmitted from the vehicle is stored while being associating with the search target.

Description

  The present invention relates to a person search system using a driver's face image taken with an in-vehicle camera.

In recent years, cameras (car-mounted cameras) that capture a driver's face image are mounted on many vehicles. This in-vehicle camera can shoot a clear face image while taking a picture using near-infrared light that is invisible to human beings, thereby suppressing the possibility of the driver feeling dazzling. Therefore, various techniques have been proposed using the features of such an in-vehicle camera.
For example, there has been proposed a technique for detecting a driver's sleepiness and a side-view driving from a driver's face image and outputting a warning sound as required (Patent Document 1). Alternatively, focusing on the fact that the driver senses the sign before the occurrence of the failure, a technique for detecting the sign of the failure before the vehicle breaks down from a slight change in the driver's facial expression (Patent Document 2), A technique (Patent Document 3) that prevents the vehicle from being stolen by determining whether or not the driver's face image matches the registered face image has been proposed.

JP 2012-84068 A JP 2011-79350 A JP 2006-306191 A

  However, the in-vehicle camera is not only a feature that a clear face image can be obtained, but also a big image that a face image can be obtained from the front direction of the driver under the same conditions in any vehicle. There is also a feature, and any of the proposed techniques has a problem that it is difficult to say that this feature is fully utilized.

  The present invention has been made in view of the above-described problems, and provides a person search system that effectively utilizes the feature of an in-vehicle camera that a clear and homogeneous face image can be obtained with any vehicle. .

  In order to solve the above-described problems, a person search system and a person search method according to the present invention use a vehicle-mounted camera provided in front of a driver seat to capture a driver's face image while driving the vehicle. Image information is generated, and the driver face image information and vehicle position information are transmitted to a database. The database stores in advance searcher face image information based on the search target person's face image. When the driver face image information matches the searcher face image information, the database is transmitted from the vehicle in association with the search target person. The vehicle position information is stored.

  The in-vehicle camera is provided in front of the driver's seat and at a relatively short distance (for example, 1 meter) from the driver's seat, so that a clear face image of the driver can be obtained and the driver's face image with high accuracy is obtained. Information can be obtained. Further, since the in-vehicle camera captures the driver's face image at the same angle and at the same distance regardless of the vehicle, it is possible to obtain uniform driver face image information. . Since the database collates the face image information with high accuracy and homogeneity in this way with the searcher face image information, the search target person can be easily found with high accuracy. Further, since the vehicle transmits not only the driver face image information but also the position information (simultaneously or separately) to the database, when the search target person is found, the position of the search target person can be specified.

  Further, in the person search system of the present invention described above, the steering angle of the steering wheel is detected, and when the steering angle is within a predetermined range including the straight traveling direction of the vehicle, the driver face image information is generated. Also good.

  If the traveling direction of the vehicle is almost straight, the driver is likely driving while looking forward. Therefore, when the steering angle of the steering wheel is within a predetermined range and the traveling direction of the vehicle is close to straight, the face image is taken to generate the driver face image information. In this way, it is possible to generate driver face image information by capturing a clear face image from the front of the driver with a high probability.

  In the person search system of the present invention described above, driver face image information may be generated when the vehicle starts to travel.

  When the vehicle starts to travel, there is a possibility that the vehicle has been started after the vehicle has been temporarily stopped and the driver has changed. Thus, since the driver image information is generated when there is a possibility that the driver has been changed, whether the driver after this change is a search target person without missing the timing at which the driver changed. Can be determined.

  In the person search system of the present invention described above, a newly generated driver face image information is generated when the driver face image information generated before the driver face image information is different from the driver face image information generated before the driver face image information. The driver's face image information thus made may be transmitted to a database.

  The newly generated driver face image information is different from the driver face image information generated before that when the driver changes. Since the driver image information is transmitted to the database when the driver changes in this way, it is possible to determine whether or not the driver after this change is a search target person at the timing when the driver changes. Become.

It is explanatory drawing which shows the structure of the person search system 1 of a present Example. It is a flowchart of the driver information transmission process which the control part 111 of a present Example performs. It is a flowchart of the search target person detection process which the database 20 of a present Example performs. It is explanatory drawing which shows notionally the memory content of the database 20 of a present Example. It is a flowchart of the driver information transmission process which the control part 111 of the modification 1 performs. 12 is a flowchart of driver information transmission processing executed by a control unit 111 according to Modification 2; It is a flowchart of the search subject person detection process which the database 20 of the modification 2 performs. 10 is a flowchart of driver information transmission processing executed by a control unit 111 according to Modification 3;

Hereinafter, examples will be described in order to clarify the contents of the present invention described above.
A. Device configuration :
FIG. 1 shows a configuration of a person search system 1 of the present embodiment. As shown in the figure, the person search system 1 according to the present embodiment includes an in-vehicle camera 100 that is mounted on a vehicle 10 and captures a driver's face image, a process that generates a facial feature amount from the driver's face image, and the like. The control device 110 that performs the above and the vehicle 10 are provided at a place different from the vehicle 10 and are configured by a database 20 that registers information on the search target person.

The in-vehicle camera 100 is an upper part of the instrument panel and is provided in front of the driver's seat and at a predetermined distance (for example, about 1 meter) from the driver's seat, and emits near-infrared light that is invisible to humans, The driver's face image is taken by detecting the reflected light. As illustrated, the control device 110 includes a control unit 111, a feature amount generation unit 112, a position information acquisition unit 113, a communication unit 114, and a storage unit 115.
Among these, the control part 111 outputs the request signal which requests | requires a face image to the vehicle-mounted camera 100 based on the state (stop / run) of the vehicle 10 detected by the vehicle speed sensor 120 with which the vehicle 10 is provided. When the feature amount generation unit 112 receives a face image photographed by the in-vehicle camera 100, the feature amount generation unit 112 generates a feature amount of the face (the shape and size of eyes, nose, mouth, etc.) from the face image, and the control unit 111 Output to.
Further, the position information acquisition unit 113 acquires the position information of the vehicle 10 based on the GPS signal received by the GPS device 130 from a GPS satellite (not shown), and outputs it to the control unit 111. The communication unit 114 receives feature amounts and position information from the control unit 111 and transmits these pieces of information to the database 20 connected via a predetermined wireless communication line. Further, the storage unit 115 stores a vehicle number of the vehicle 10 (information that can identify the vehicle 10), and the vehicle number is also transmitted to the database 20 together with the feature amount and the position information.

  On the other hand, the database 20 is obtained from the facial images of persons who need to search for a place (search target person) such as a criminal who is seeking a wanted person or a missing person from whom a search request has been issued, and the facial images. Feature values are registered in advance. Then, when the database 20 receives the feature amount of the driver from the vehicle 10, the driver of the vehicle 10 can check the feature amount with the feature amount of the search target person registered in the database 20. It is judged whether it corresponds to a person.

B. Driver information transmission process:
FIG. 2 shows a flowchart of driver information transmission processing executed by the control unit 111 of this embodiment. When the engine of the vehicle 10 is activated, the control unit 111 starts a driver information transmission process and determines whether the vehicle 10 has started running (S100). Whether or not the vehicle 10 has started traveling is determined by the control unit 111 detecting that the speed detected by the vehicle speed sensor 120 is not zero. As a result, if the vehicle 10 has not started running (S100: no), this determination process is repeated, and the vehicle 10 is in a standby state until the vehicle 10 starts running.

  When the vehicle 10 starts running (S100: yes), the control unit 111 outputs a request signal for requesting a face image to the in-vehicle camera 100, and causes the in-vehicle camera 100 to capture the driver's face image (S102). ). The in-vehicle camera 100 can shoot a clear face image even at night while suppressing the possibility that the driver feels dazzling by photographing using near-infrared light invisible to humans.

  Then, the face image captured by the in-vehicle camera 100 is output to the feature value generation unit 112, and the feature value generation unit 112 generates a face feature value from the face image (S104). It should be noted that a known method used when performing so-called face recognition can be used when generating the feature amount. For example, feature points of facial parts such as the corners of the eyes and the eyes are extracted from the captured face image, and features such as the shape and size of the facial parts (eyes, nose, mouth, etc.) based on those feature points Generate quantity.

  When the feature amount of the driver's face is generated in this way (S104), the control unit 111 reads the vehicle number of the vehicle 10 from the storage unit 115, and transmits it from the communication unit 114 to the database 20 together with the generated feature amount (S106). Since the vehicle number is information unique to the vehicle 10, the database 20 that has received the feature amount and the vehicle number can identify the vehicle 10 on which the driver having the feature amount is aboard.

  Subsequently, the control unit 111 determines whether or not a coincidence signal is received from the database 20 within a first predetermined time (for example, 15 seconds) after transmitting the feature amount and the vehicle number (S108). Here, as will be described in detail later, when the database 20 receives a feature quantity from the vehicle 10, whether or not there is a search target person having the feature quantity, that is, the driver of the vehicle 10 is the search target person. Determine whether or not. As a result, when it is determined that the driver of the vehicle 10 is a search target person (when the received feature amount matches the search target person), the database 20 transmits a match signal indicating that to the vehicle 10. . Therefore, in the process of S108 described above, the control unit 111 of the vehicle 10 determines whether or not the communication unit 114 has received the coincidence signal within the first predetermined time after transmitting the feature amount and the vehicle number ( S108).

  As a result, when the communication unit 114 receives the coincidence signal (S108: yes), that is, when the database 20 determines that the driver is a search target person, the position information acquisition unit 113 displays the driver's whereabouts. Is acquired from the GPS device 130 (S110). Subsequently, the control unit 111 transmits the vehicle number of the vehicle 10 together with the position information from the communication unit 114 to the database 20 (S112). That is, the position information and the vehicle number of the vehicle 10 that the search target person is driving are transmitted to the database 20 by performing the processes of S110 and S112. And the process which transmits these information is repeated whenever the 2nd predetermined time (for example, 10 second) passes (S114). For this reason, even if the position of the vehicle 10 that the search target person is driving moves, the database 20 can acquire the latest position of the search target person.

On the other hand, if it is determined in the process of S108 that the coincidence signal has not been received within the first predetermined time after transmitting the feature amount and the vehicle number (S108: no), the driver of the vehicle 10 It is considered that the database 20 has determined that the search target person is not a search target person (if the feature amount transmitted from the vehicle 10 does not match the search target person). In this case, returning to the process of S100 at the head, when the vehicle 10 starts running again, a series of processes after S102 are repeated.
Here, the driver of the vehicle 10 may change driving to another person while the engine is activated, and it is considered that the vehicle 10 is temporarily stopped at the time of the change. For this reason, as described above, when the vehicle 10 starts running (from a stopped state) (S100: yes), if a series of processes after S102 is repeated, the engine of the vehicle 10 is activated. Even if the driver changes as it is, it is possible to determine whether or not the driver after the change is a search target person.

  As described above, when the vehicle 10 according to the present embodiment receives the coincidence signal after transmitting the feature amount, the driver having the feature amount is the search target person. The position information and the vehicle number of the vehicle 10 are periodically transmitted to the database 20 in order to track the person.

C. Search target person detection processing:
FIG. 3 shows a flowchart of the search target person detection process executed by the database 20. The database 20 determines whether or not the driver of the vehicle 10 is a search target person by performing the following processing in accordance with various information transmitted from the vehicle 10, and the driver If it is a search target person, the position is detected.

When the search target person detection process is started, the database 20 first determines whether or not the feature amount and the vehicle number have been received from the vehicle 10 (S200). As described above with reference to FIG. 2, the feature amount and the vehicle number are transmitted from the vehicle 10 when the vehicle 10 starts traveling.
As a result of the determination process in S200, if the feature amount and the vehicle number have not been received (S200: no), it is determined whether or not the position information and the vehicle number have been received from the vehicle 10 (S202). As described above with reference to FIG. 2, the location information and the vehicle number are transmitted from the vehicle 10 when the database 21 determines that the driver of the vehicle 10 is a search target person (a coincidence signal has been received). The position information and the vehicle number information are repeatedly transmitted from the vehicle 10 in order to track the vehicle 10 driven by the search target person.
If the position information and the vehicle number are not received as a result of the determination process in S202 (S202: no), the process returns to the first S200 process.

On the other hand, when the feature amount and the vehicle number are received from the vehicle 10 (S200: yes), the received feature amount is collated with the feature amount of the search target person registered in the database 20 (S204).
FIG. 4 conceptually shows information stored in the database 20. As illustrated, the search target person, the file name of the face image of the search target person (face image file), and the feature amount extracted from the face image are registered in advance in the database 20. Further, it is possible to store a vehicle number and position information in association with each search target person.
The database 20 collates the feature amount received from the vehicle 10 with the feature amount of the search target person registered in advance in the process of S204. As a result, when the feature quantity received from the vehicle 10 matches the feature quantity of any search target person (S206: yes), the vehicle number is stored in association with the matched search target person (S208). ). Thereby, the vehicle number of the vehicle 10 driven by the search target person is stored. In the example shown in FIG. 4, the feature number received from the vehicle 10 matches the feature amount of the search target person “C”, so that the vehicle number “nnnnn” of the vehicle 10 is associated with the search target person “C”. It is remembered.

  Subsequently, the database 20 transmits a coincidence signal indicating that the driver of the vehicle 10 is a search target person to the vehicle 10 (S210). When the coincidence signal is transmitted in this way, the database 20 returns to the process of S200 at the head and repeats the series of processes described above. Even when it is determined in S206 that the driver of the vehicle 10 is not a search target person (S206: no), the process returns to the top S200 as it is, and the series of processes described above is repeated.

Here, the vehicle 10 that has received the coincidence signal from the database 20, that is, the vehicle 10 that is determined to be driven by the search target person, transmits the position information and the vehicle number to the database 20 as described above (FIG. 2). Of S112). Therefore, when the database 20 receives the position information and the vehicle number (S202: yes), the database 20 identifies the search target person whose vehicle number has already been stored in the process of S208 described above, and corresponds to the search target person. In addition, the positional information is stored (S212). Thereby, the position information of the search target person (vehicle 10 driven by) is stored. In the example shown in FIG. 4, the position information “La, Lo” is stored in association with the search target person “C” by receiving the vehicle number of the search target person “C”.
As described above with reference to FIG. 2, the position information and the vehicle number are transmitted from the vehicle 10 to the database 20 every second predetermined time (see S114 in FIG. 2), and thus stored in the database 20. The position information is also updated every second predetermined time.

  When the position information of the search target person is stored in this way (S212), the position information and the vehicle number together with the identification information (name, etc.) of the search target person are notified to the searcher (search person in charge) (S214). ). In addition, as a person in charge of searching, when a person to be searched is a criminal or a missing person, a staff member of a police station (related organization) searching for these persons can be exemplified. Of course, the person in charge of the search finds out the person to be searched for based on the position information of the vehicle 10 and the vehicle number.

  As described above, in the person search system 1 of the present embodiment, a face image is displayed from the front of the driver while the vehicle 10 is traveling with the in-vehicle camera 100 provided in front of the driver's seat and within a predetermined distance from the driver's seat. A feature is generated from the face image and transmitted to the database 20. The database 20 stores the feature amount of the search target person in advance, and if the feature amount of the driver transmitted from the vehicle 10 matches the feature amount of the search target person, the driver of the vehicle 10 The position information transmitted from the vehicle 10 is stored as the position information of the search target person.

  Conventionally, in such a system for searching for a person, a face image photographed under various conditions is used. For example, a face image photographed from a horizontal direction or a face image photographed from a distance is used. It was used. Face images shot under such various conditions need to be subjected to heavy correction processing when the feature amount is collated with the feature amount of the search target person, or an accurate feature amount can be extracted. Since it was not possible, it was difficult to find a search target person.

On the other hand, in the person search system 1 of the present embodiment, the in-vehicle camera 100 is provided in front of the driver's seat and at a predetermined distance (for example, about 1 meter) from the driver's seat. A clear face image can be obtained.
Further, in most vehicles, the in-vehicle camera 100 is provided in front of the driver's seat and at a short distance from the driver's seat. A face image can be obtained. Since the feature amount is generated from such a clear and homogeneous face image and the feature amount is collated with the feature amount of the search target person, the search target person can be easily found with high accuracy. When the search target person is found, the vehicle 10 driven by the search target person transmits the position information of the vehicle 10 to the database 20, so that the position of the search target person can be specified.

Note that the control unit 111 and the feature value generation unit 112 according to the present embodiment capture the driver's feature value (driver face) by capturing the driver's face image with the in-vehicle camera 100 when the vehicle 10 starts to travel. Image information). Therefore, the control unit 111 and the feature amount generation unit 112 of this embodiment correspond to the “generation unit” in the present invention. Further, since the position information acquisition unit 113 of the present embodiment acquires the position information of the vehicle 10, it corresponds to the “position information acquisition means” in the present invention. In addition, the communication unit 114 according to the present embodiment transmits the driver's feature amount (driver face image information) and position information to the database, and thus corresponds to the “transmission unit” in the present invention.
Further, the database according to the present embodiment stores a feature amount (searcher face image information) based on the face image of the search target person. Further, the feature amount (driver face image information) of the driver transmitted by the communication unit 114 (transmission means) of the vehicle 10 is collated with the feature amount (searcher face image information) of the search target person. Further, as a result of the collation, if the driver's feature amount (driver face image information) matches the search subject's feature amount (searcher face image information), the communication unit 114 ( The position information transmitted by the transmission means) is stored. Therefore, the database of the present embodiment corresponds to “search target person storage means”, “collation means”, and “position information storage means” in the present invention.

  In the above-described embodiment, the driver's feature amount is described as being transmitted from the communication unit 114 of the vehicle 10. However, instead of the feature amount, the driver's face image may be transmitted. And the database 20 is good also as producing | generating the feature-value from the received face image, and collating the feature-value with the search subject's feature-value. In this way, the control device 110 of the vehicle 10 does not need to generate a feature amount, and thus the processing load on the control device 110 can be reduced.

D. Variations There are several variations in the person search system 1 of the embodiment described above. In the following, these modified examples will be described focusing on the differences from the embodiments. In the modification described below, the same components as those in the above-described embodiment are denoted by the same reference numerals, and detailed description thereof is omitted.

D-1. Modification 1
In the above-described embodiment, every time the vehicle 10 starts traveling until the vehicle 10 receives the coincidence signal (until the database 20 determines that the driver of the vehicle 10 is a search target person), Since the driver may have changed, the feature amount of the driver has been described as being transmitted to the database 20.
However, the vehicle 10 often starts running without the driver changing, such as when the vehicle 10 is temporarily stopped based on a traffic light or a sign. Accordingly, it is possible to determine whether or not the driver has been changed in the vehicle 10 and to transmit the feature amount of the driver to the database 20 when it is determined that the driver has changed.

  FIG. 5 shows a flowchart of the driver information transmission process of the first modification. When the driver information transmission process of the first modification is started, the control unit 111 starts the driver information transmission process when the engine is started, based on the vehicle speed detected by the vehicle speed sensor 120, as in the embodiment. It is determined whether or not 10 has started running (S300). When the vehicle 10 starts running (S300: yes), the in-vehicle camera 100 captures the driver's face image (S302), and causes the feature amount generation unit 112 to generate a feature amount from the face image (S304).

Subsequently, in the driver information transmission process of the first modification, the feature amount generated from the driver's face image is stored at a predetermined address in the storage unit 115 (S306). Then, the control unit 111 determines whether or not the feature amount generated this time matches the feature amount generated previously (S308). That is, also in the first modification, as in the above-described embodiment, until the coincidence signal is received, the driver's face image is photographed and a feature amount is generated from the face image.
Therefore, in the processing of S308, the feature amount generated immediately before (the feature amount generated from the face image captured immediately before) is read from the storage unit 115, and such feature amount and the newly generated feature this time are included. It is determined whether or not the amounts match (S308).

As a result, when it is determined that the feature amount generated this time does not match the feature amount generated immediately before (S308: no), the driver having the feature amount generated this time is generated immediately before. Since it is different from the driver having the feature amount, it can be determined that the driver has changed. Therefore, in this case, the control unit 111 transmits the feature amount generated this time together with the vehicle number of the vehicle 10 from the communication unit 114 to the database 20 (S310). Thereby, it is possible to determine whether or not the driver after the change is a search target person at the timing of the change of the driver.
Further, when it is not determined that the driver has changed, the feature amount and the vehicle number are not transmitted to the database 20, so that the amount of information transmitted from the vehicle 10 to the database 20 can be suppressed.

  After the feature amount is transmitted to the database 20 in this manner (S310), it is determined whether or not the coincidence signal is received within the first predetermined time after the feature amount and the vehicle number are transmitted as in the above-described embodiment. (S312) If the coincidence signal is received within the first predetermined time (S312: yes), it is determined by the database 20 that the driver of the vehicle 10 is the search target person. And the process of transmitting the position information and the vehicle number to the database 20 is repeated each time the second predetermined time elapses (S314 to S318).

  Here, when the engine is started (or stopped), the control unit 111 initializes (clears) the feature amount stored in the storage unit 115. Therefore, when the feature amount generated this time is generated from the face image first taken after the engine is started, the feature amount that can be compared in the determination process of S308 described above (generated immediately before). Feature amount) does not exist. In this case, of course, the feature quantity generated this time (the feature quantity generated first after the engine is started) is transmitted to the database 20 together with the vehicle number (S310).

  Note that the communication unit 114 according to the first modification includes the newly generated driver feature amount (driver face image information) and the driver feature generated before the feature amount (the driver face image information). When the amount (driver face image information) is different, the newly generated feature amount (driver face image information) of the driver is transmitted to the database 20. Therefore, the communication unit 114 of the first modification corresponds to “communication means” in the present invention.

D-2. Modification 2
In the embodiment described above, the search target person registered in advance in the database 20 is searched. On the other hand, even if it is not registered in the database 20 in advance, when it is necessary to search for the driver of the vehicle 10 due to traffic violation or the like, the driver may be searched.

  FIG. 6 shows a flowchart of driver information transmission processing according to the second modification. When the driver information transmission process of Modification 2 is started, the control unit 111 determines whether or not the vehicle speed of the vehicle 10 exceeds the legal speed (S400). Whether or not the vehicle 10 exceeds the legal speed is determined by the control unit 111 detecting that the speed detected by the vehicle speed sensor 120 exceeds the legal speed stored in the storage unit 115. As a result, if the legal speed is not exceeded (S400: no), this determination process is repeated.

On the other hand, when the vehicle speed of the vehicle 10 exceeds the legal speed (speed violation) (S400: yes), the following processing is performed to start the search for the driver of the vehicle 10. First, the control unit 111 outputs a request signal for requesting a face image to the in-vehicle camera 100, and causes the in-vehicle camera 100 to capture the driver's face image (S402).
Then, when receiving the face image from the in-vehicle camera 100, the feature value generation unit 112 of Modification 2 outputs the face image (driver face image information) as it is to the control unit 111. As will be described later, this face image is a clue when finding out the driver of the vehicle 10 that has violated the speed.

  Subsequently, the position information acquisition unit 113 is caused to acquire the position information of the vehicle 10 from the GPS device 130 (S404). And a driver | operator's face image is transmitted to the database 20 from the communication part 114 with the positional information and vehicle number of the vehicle 10 (S406).

  As described above, in the second modification, when the vehicle speed of the vehicle 10 exceeds the legal speed (speed violation), the face image of the driver of the vehicle 10 is transmitted to the database 20. When the database 20 receives the face image of the driver of the vehicle 10, the database 20 performs processing for searching for the driver regardless of whether or not the driver is registered (stored) in advance.

FIG. 7 shows a flowchart of search target person detection processing according to the second modification. When the search target person detection process according to the second modification is started, the database 20 determines whether or not a face image, position information, and a vehicle number have been received (S500). As a result, if these pieces of information have not been received (S500: no), this determination process is repeated, and a standby state is entered until reception.
On the other hand, when such information is received (S500: yes), a facial feature amount is generated from the received face image (S502). Then, the generated feature amount is collated with the feature amount of the search target person registered (stored) in advance (S504). As a result, when the generated feature amount matches the feature amount of the search target person, that is, when it is determined that the driver of the vehicle 10 is the search target person registered (stored) in advance (S506: yes). Information related to the search target (related information), position information, and vehicle number registered in advance in the database 20 is notified to the related organizations such as the police station (S508). A staff member of the related organization that conducts the search can find out the driver of the vehicle 10 using these pieces of information as clues.

  On the other hand, when it is determined in the determination process of S506 that the driver of the vehicle 10 is not a search target person (S506: no), the face image received from the vehicle 10 is displayed together with the position information and the vehicle number at the police station. Etc. (S510). Thereby, even if the driver of the vehicle 10 is not a search target person registered in advance in the database 20, a member of the related organization that performs the search uses the face image, position information, and vehicle number received by the database 20 from the vehicle 10. Using the clue, the driver of the vehicle 10 can be found.

  In the second modification described above, the case where the driver has made a traffic violation has been described as an example of the condition for determining that the driver needs to search. However, the present invention is not limited to such a case. For example, when it is determined that the driver is a criminal (criminal) such as when the driver steals the vehicle 10 and drives the vehicle, It is good also as searching for a driver.

Here, whether or not the driver of the vehicle 10 is a criminal may be determined as follows. A feature amount generated from a face image of the owner of the vehicle 10 or the family of the owner is stored in the storage unit 115 in advance. And if the control part 111 judges that the vehicle 10 started driving | running | working, it will make the vehicle-mounted camera 100 image | photograph a driver | operator's face image. Subsequently, a feature amount is generated from the driver's face image, and whether or not the generated feature amount (driver's feature amount) matches the candidate feature amount stored in the storage unit 115. Determine whether.
As a result, when the feature amount of the driver matches the feature amount of the candidate, the driver is considered to be a regular driver (candidate). On the other hand, when the feature amount of the driver does not match the feature amount of the candidate, the driver can be considered as a person (suspicious person) who is not a regular driver (candidate). In this case, the control unit 111 transmits the driver's face image captured by the in-vehicle camera 100 together with the position information and the vehicle number of the vehicle 10 from the communication unit 114 to the database 20 in the same manner as the process of S406 described above. . By doing in this way, when the driver is not a regular driver, the staff of the related organization that performs the search uses the face image, the position information, and the vehicle number received by the database 20 from the vehicle 10 as clues. Can find a driver.

D-3. Modification 3
In the above-described embodiments, the process of capturing the driver's face image is described as being performed while the vehicle 10 is traveling. Not limited to this, the process of capturing the driver's face image may be performed when the vehicle 10 is traveling and the steering angle of the steering wheel is within a predetermined range.

  FIG. 8 shows a flowchart of the driver information transmission process of the third modification. Also in the third modification, the control unit 111 determines whether or not the vehicle 10 has started traveling based on the speed detected by the vehicle speed sensor 120 (S600). As a result, if the vehicle 10 has not started running (S600: no), this determination process is repeated, and the vehicle 10 enters a standby state until the vehicle 10 starts running.

When the vehicle 10 starts traveling (S600: yes), the control unit 111 determines that the steering angle of the steering wheel is within a predetermined range including the straight traveling direction of the vehicle 10 (for example, −10 degrees with the straight traveling direction of the vehicle 10 being 0 degrees). It is judged whether it is within the range of ˜ + 10 degrees (S602). The steering angle of the steering wheel is detected by a steering angle sensor (not shown) provided in the vehicle 10 and output to the control unit 111.
As a result, when the steering angle of the steering wheel is within a predetermined range (S602: yes), a request signal for requesting a face image is output to the in-vehicle camera 100 to cause the in-vehicle camera 100 to capture the driver's face image ( S604). Then, the feature amount generation unit 112 generates a feature amount from the photographed face image (S606).

Here, if the traveling direction of the vehicle 10 is almost straight, the driver is highly likely to drive while looking forward. Therefore, in Modification 3, when the steering angle of the steering wheel is within a predetermined range and the traveling direction of the vehicle is in a state close to straight traveling, a face image is captured to generate a feature amount. In this way, it is possible to generate a feature amount by capturing a clear face image from the front of the driver with a high probability.
Thereafter, the process proceeds to the process of S106 described above with reference to FIG. 2 (process for transmitting the feature amount and the vehicle number to the database 20), and the subsequent series of processes is executed as in the case of the above-described embodiment. To do.

On the other hand, when it is determined in the determination process of S602 that the steering angle of the steering wheel is not within the predetermined range (S602: no), the steering wheel is largely turned to either the left or right, so that the driver Is likely to drive while looking at the left or right direction. In this case, based on the vehicle speed detected by the vehicle speed sensor 120, it is determined whether or not the vehicle 10 is traveling (S608).
As a result, if it is determined that the vehicle 10 is traveling (S608: yes), the process returns to the above-described process of S602 again to determine whether or not the steering angle of the steering wheel is within a predetermined range (S602). . In this way, when the steering angle of the steering wheel is within a predetermined range while the vehicle 10 is traveling (when the vehicle 10 is in a state of being nearly straight), a face image of the driver is taken (S604).
If it is determined in S608 that the vehicle 10 is not traveling (S608: no), that is, if the vehicle 10 has stopped traveling without being in a state of being nearly straight, the first S600 processing is performed. When the vehicle 10 starts traveling again (S600: yes), the above-described processing from S602 onward is repeated.

  The steering angle sensor of the third modification corresponds to the “steering angle detecting means” in the present invention because it detects the steering angle of the steering wheel of the vehicle 10. In addition, the control unit 111 and the feature amount generation unit 112 of Modification 3 generate the feature amount (driver face image information) of the driver when the steering angle is within a predetermined range including the straight traveling direction of the vehicle 10. . Therefore, the control unit 111 and the feature value generation unit 112 of the third modification correspond to “generation means” in the present invention.

  As mentioned above, although the person search system of an Example and a modification was demonstrated, this invention is not limited to said Example and a modification, It can implement in a various aspect in the range which does not deviate from the summary.

1 ... person search system, 20 ... database, 10 ... vehicle,
110: Control device, 111: Control unit, 112 ... Feature quantity generation unit,
113 ... Location information acquisition unit, 114 ... Communication unit, 115 ... Storage unit,
120: Vehicle speed sensor, 130: GPS device.

Claims (5)

A person search system comprising a vehicle and a database connected to the vehicle via a predetermined communication line,
The vehicle is
An in-vehicle camera installed in front of the driver's seat;
Generating means for generating driver face image information by photographing a driver's face image with the in-vehicle camera;
Position information acquisition means for acquiring position information of the vehicle;
Transmitting means for transmitting the driver face image information and the position information to the database;
The database is
Search target person storage means for storing searcher face image information based on the search target person's face image;
Collating means for collating the driver face image information transmitted by the transmitting means with the searcher face image information;
Position information storage means for storing the position information transmitted by the transmission means in association with the search target person when the driver face image information matches the searcher face image information as a result of the collation by the collation means A person search system comprising and. The person search system according to claim 1,
The vehicle includes steering angle detection means for detecting a steering angle of a steering wheel of the vehicle,
The generation means is a person search system which is means for generating the driver face image information when the steering angle is within a predetermined range including a straight traveling direction of the vehicle. The person search system according to claim 1 or 2,
The generation means is a person search system which is means for generating the driver face image information when the vehicle starts to travel. The person search system according to any one of claims 1 to 3,
The transmission means generates the newly generated driver face when the newly generated driver face image information is different from the driver face image information generated before the driver face image information. A person search system which is means for transmitting image information to the database. A person search method using a vehicle and a database connected to the vehicle via a predetermined communication line,
In the vehicle, generating driver face image information by photographing a driver's face image with an in-vehicle camera provided in front of the driver's seat;
Obtaining the position information of the vehicle in the vehicle;
Transmitting the driver face image information and the position information from the vehicle to the database;
Collating the driver face image information transmitted from the vehicle with searcher face image information based on the face image of the search target person in the database;
A person search method comprising: storing the position information in association with the searcher face image information when the driver face image information matches the searcher face image information in the database.

Images