JP2009019981A - Object position detecting device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an object position detecting device monitoring a large spatial living room or the like at lower cost than before. <P>SOLUTION: In an object position detecting device 1, a pick-up image data receiving part 36 receives pick-up image data from a swing imaging device 2. A specified pixel coordinate deriving means 34d derives specified pixel coordinates. A specified information table extracting means extracts a specified information table Tm1 by collating a combination of yaw angle and pitch angle, or yaw angle at a lens part 22 of the swing imaging device with an information storing part 34. A first pole coordinate deriving means 34e derives the first pole coordinate by utilizing the specified pixel coordinates and the specified information table. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to an object position detection apparatus capable of detecting the position of an object, particularly a moving body such as a person.

In the past, “means for detecting the head of a person in the image data captured by the monitoring camera, and coordinates on the image data of the head (hereinafter referred to as“ first coordinates ”) in the real space of the imaging range of the monitoring camera. Has been proposed that includes a conversion table corresponding to the coordinates of the floor portion (hereinafter referred to as “second coordinates”) and means for deriving the second coordinates by comparing the first coordinates with the conversion table ”. (For example, refer to Patent Document 1). If such a monitoring camera system is used, the position of a person can be specified accurately.
JP 2003-189295 A

  However, in such a surveillance camera system, a large number of surveillance cameras are required to monitor large space rooms such as offices. Therefore, it is very expensive to introduce such a monitoring camera system.

  An object of the present invention is to provide an object position detection device capable of monitoring a large space room or the like at a lower cost than before.

  An object position detection apparatus according to a first aspect of the present invention includes a swing imaging machine, an information storage unit, a captured image data receiving unit, a specific pixel coordinate deriving unit, a specific information table extracting unit, and a first polar coordinate deriving unit. The head swing imaging device has a lens unit and a turning drive mechanism. The turning drive mechanism can turn the lens unit in the yaw direction and the pitch direction. The swing imaging device may be a monocular imaging device or a compound eye imaging device, but a monocular imaging device is preferable from the viewpoint of cost. An information table is stored in the information storage unit. The information table is created for each combination of the yaw angle and pitch angle of the lens unit or for each pitch angle. In this information table, the first polar coordinates are assigned to the pixel coordinates of the swing imaging device. For the distance component of the first polar coordinate, “the distance to the foot calculated in advance by taking into account the average height (the distance from the actual polar coordinate, which varies depending on the distance from the origin) has already been added or subtracted. Is registered). Further, only the portion near the center of the first polar coordinate may be replaced with the orthogonal coordinate. Further, when the first polar coordinates are the actual polar coordinates of the floor surface, the assignment to the pixel coordinates may be determined by calculation or may be determined by edge extraction by an image processing technique. The captured image data receiving unit receives captured image data from the swing imaging device. The specific pixel coordinate deriving unit derives specific pixel coordinates. Here, the “specific pixel coordinates” are pixel coordinates of a specific part of the captured image data. Further, the “specific part” mentioned here is, for example, a part of a specific color. As the “specific pixel coordinate deriving means”, a known method can be used. The specific information table extracting means collates the combination of the yaw angle and the pitch angle or the pitch angle with the information storage unit and extracts the specific information table. The “specific information table” here is a specific information table. The first polar coordinate deriving means derives the first polar coordinate using the specific pixel coordinates and the specific information table.

  In this object position detection device, a swing imaging device is employed as a monitoring camera. For this reason, in this object position detection device, even when a large space room such as an office is to be monitored, there may be one or several monitoring cameras (when there is a shield in the office space or the like). That's enough. Therefore, this object position detection apparatus can be introduced at a lower cost than in the past.

  An object position detection apparatus according to a second aspect is the object position detection apparatus according to the first aspect, wherein the first polar coordinate deriving means derives the first polar coordinate by comparing specific pixel coordinates with a specific information table.

  In the object position detection device, the first polar coordinate deriving unit derives the first polar coordinate by collating the specific pixel coordinate with the specific information table. For this reason, in this object position detection apparatus, the first polar coordinates are quickly derived.

  An object position detection device according to a third aspect of the invention is the object position detection device according to the first or second aspect of the invention, and the information table is created for each pitch angle. In this information table, the first polar coordinates are assigned to the pixel coordinates of the swing imaging device. The specific information table extracting means extracts the specific information table by comparing the pitch angle with the information storage unit. The object position detection device further includes second polar coordinate deriving means. The second polar coordinate deriving unit derives the second polar coordinate by adding or subtracting the yaw angle to or from the angle component of the first polar coordinate derived by the first polar coordinate deriving unit.

  In this object position detection device, an information table is created for each pitch angle. Then, the second polar coordinate deriving unit derives the second polar coordinate by adding or subtracting the yaw angle to the angle component of the first polar coordinate derived by the first polar coordinate deriving unit. For this reason, in this object position detection apparatus, the number of information tables that should be stored in the information storage unit can be reduced. Therefore, in this object position detection device, the amount of information to be stored can be reduced. Therefore, this object position detection apparatus can be introduced at a lower cost.

  An object position detection device according to a fourth aspect of the present invention is the object position detection device according to the first or second aspect of the present invention, wherein the pixel coordinates are orthogonal coordinates with the central portion in the yaw direction being 0, 1 or −1. is there. The information table is created for each combination of the yaw angle and pitch angle of the lens unit or for each pitch angle. In this information table, the first polar coordinates are assigned only to the half-pixel coordinates among the pixel coordinates of the swing imaging device. Here, the “half pixel coordinates” are pixel coordinates that are half of the yaw direction (including 0 when the center in the yaw direction is 0). The object position detection device further includes first code conversion means. The first code conversion means performs code conversion of the yaw direction component of the specific pixel coordinate when the specific pixel coordinate does not match any of the pixel coordinates of the half pixel coordinate. The first polar coordinate deriving unit collates the specific pixel coordinate whose yaw direction component is code-converted by the first code converting unit with the specific information table when the specific pixel coordinate does not coincide with any of the half-pixel coordinates. To derive the first polar coordinate. The object position detection device further includes second code conversion means. The second code converting means converts the angle component of the first polar coordinate derived by the first polar coordinate deriving means when the specific pixel coordinate does not match any of the half-pixel coordinates.

  In this object position detection device, the first polar coordinates are assigned only to the half pixel coordinates among the pixel coordinates of the swing imaging device in the information table. For this reason, in this object position detection apparatus, the number of polar coordinates assigned to pixel coordinates can be reduced. Therefore, in this object position detection device, the amount of information to be stored can be reduced. Therefore, this object position detection apparatus can be introduced at a lower cost.

  An object position detection device according to a fifth aspect of the present invention is the object position detection device according to the first or second aspect of the present invention, wherein the pixel coordinates are orthogonal coordinates in which the center of the yaw direction is 0, 1 or −1. is there. The information table is created for each combination of the yaw angle and pitch angle of the lens unit or for each pitch angle. Further, in this information table, the first polar coordinates are assigned only to the pixel coordinates on the plus side in the yaw direction (including 0 when the central portion in the yaw direction is 0) among the pixel coordinates of the swing imaging device. The object position detection device further includes absolute value converting means. The absolute value converting means converts the yaw direction component of the specific pixel coordinate into an absolute value. The first polar coordinate deriving unit derives the first polar coordinate by collating the specific pixel coordinates whose absolute value is the yaw direction component by the absolute value converting unit with the specific information table. The object position detection device further includes a yaw direction component code determination unit and a third code conversion unit. The yaw direction component code determination means determines whether or not the yaw direction component of the specific pixel coordinate is a negative value. The third code conversion unit codes the angle component of the first polar coordinate derived by the first polar coordinate deriving unit when the yaw direction component of the specific pixel coordinate is determined to be a negative value by the yaw direction component code determination unit. Convert.

  In this object position detection device, the first polar coordinates are assigned only to the pixel coordinates on the plus side in the yaw direction (including 0 when the center in the yaw direction is 0) among the pixel coordinates of the swing imaging device in the information table. ing. For this reason, in this object position detection apparatus, the number of polar coordinates assigned to pixel coordinates can be reduced. Therefore, in this object position detection device, the amount of information to be stored can be reduced. Therefore, this object position detection apparatus can be introduced at a lower cost.

  The object position detection device according to a sixth aspect of the invention is the object position detection device according to the fourth or fifth aspect of the invention, and the information table is created for each pitch angle. The specific information table extracting means extracts the specific information table by comparing the pitch angle with the information storage unit. The object position detection device further includes third polar coordinate deriving means. The third polar coordinate deriving unit derives the third polar coordinate by adding or subtracting the yaw angle to or from the angle component of the first polar coordinate converted by the second code converting unit or the third code converting unit.

  In this object position detection device, an information table is created for each pitch angle. Then, the third polar coordinate deriving unit derives the third polar coordinate by adding or subtracting the yaw angle to or from the angle component of the first polar coordinate converted by the second code converting unit or the third code converting unit. For this reason, in this object position detection apparatus, the number of information tables that should be stored in the information storage unit can be reduced. Therefore, in this object position detection device, the amount of information to be stored can be reduced. Therefore, this object position detection apparatus can be introduced at a lower cost.

  An object position detection apparatus according to a seventh aspect includes a fish-eye imager, an information storage unit, a captured image data receiving unit, a specific pixel coordinate deriving unit, and an eleventh polar coordinate deriving unit. An information table is stored in the information storage unit. In this information table, eleventh polar coordinates are assigned to the pixel coordinates of the fish-eye imager. The distance component of the eleventh polar coordinate is “a distance that takes into account the average height in advance (a value obtained by adding or subtracting a predetermined distance (which varies depending on the distance from the origin) from the actual polar coordinate distance)”. be registered. The captured image data receiving unit receives captured image data from the fisheye imager. The specific pixel coordinate deriving unit derives specific pixel coordinates. Here, the “specific pixel coordinates” are pixel coordinates of a specific part of the captured image data. The eleventh polar coordinate deriving means derives the eleventh polar coordinate by collating the specific pixel coordinate with the information table.

  In this object position detection device, a fish-eye imager is employed as a monitoring camera. For this reason, in this object position detection device, even when a large space room such as an office is to be monitored, there may be one or several monitoring cameras (when there is a shield in the office space or the like). That's enough. Therefore, this object position detection apparatus can be introduced at a lower cost than in the past.

  An object position detection apparatus according to an eighth aspect of the present invention is the object position detection apparatus according to the seventh aspect of the present invention, wherein the pixel coordinates are orthogonal coordinates with the central part in the horizontal direction or the vertical direction being 0, 1 or −1. . In the information table, the eleventh polar coordinate is assigned only to the half pixel coordinate among the pixel coordinates of the fisheye imager. Here, the “half pixel coordinates” are pixel coordinates of a half in the horizontal direction or the vertical direction (including 0 when the center in the horizontal direction or the vertical direction is 0). The object position detection device further includes eleventh code conversion means. The eleventh code conversion means converts the horizontal component or the vertical component of the specific pixel coordinate when the specific pixel coordinate does not match any of the pixel coordinates of the half pixel coordinate. The eleventh polar coordinate deriving means stores, in the information table, the specific pixel coordinates obtained by code-converting the horizontal component or the vertical component by the eleventh code converting means when the specific pixel coordinates do not coincide with any one of the half-pixel coordinates. The eleventh polar coordinate is derived by collation. The object position detection device further includes a twelfth code conversion unit. The twelfth code converting means converts the angle component of the first polar coordinate derived by the eleventh polar coordinate deriving means when the specific pixel coordinate does not match any of the pixel coordinates of the half pixel coordinates.

  In this object position detection apparatus, the eleventh polar coordinate is assigned only to the half pixel coordinate among the pixel coordinates of the fisheye imager in the information table. For this reason, in this object position detection apparatus, the number of polar coordinates assigned to pixel coordinates can be reduced. Therefore, in this object position detection device, the amount of information to be stored can be reduced. Therefore, this object position detection apparatus can be introduced at a lower cost.

  An object position detection device according to a ninth aspect is the object position detection device according to the seventh aspect, wherein the pixel coordinates are orthogonal coordinates with the center portion in the horizontal direction or the vertical direction being 0, 1 or −1. . In the information table, the eleventh polar coordinate is assigned only to the pixel coordinate on the plus side in the horizontal direction or the vertical direction (including 0 when the center in the horizontal direction or the vertical direction is 0) among the pixel coordinates of the fisheye imager. . The object position detection device further includes absolute value converting means. The absolute value converting means converts the horizontal component or vertical component of the specific pixel coordinate into an absolute value. The eleventh polar coordinate deriving means derives the eleventh polar coordinate by collating the specific pixel coordinates whose horizontal component or vertical component is absolute valued by the absolute value converting means with the information table. The object position detection apparatus further includes a horizontal component equality code determination unit and a thirteenth code conversion unit. The horizontal component etc. sign determination means determines whether the horizontal component or the vertical component of the specific pixel coordinate is a negative value. The thirteenth code conversion means determines the angle component of the eleventh polar coordinate derived by the eleventh polar coordinate derivation means when the horizontal or vertical direction of the specific pixel coordinate is determined to be a negative value by the horizontal component etc. sign determination means. Is converted.

  In this object position detection apparatus, only the pixel coordinates on the plus side in the horizontal direction or the vertical direction among the pixel coordinates of the swing image pickup device in the information table (including 0 when the center in the horizontal direction or the vertical direction is 0) are included. An eleventh polar coordinate is assigned. For this reason, in this object position detection apparatus, the number of polar coordinates assigned to pixel coordinates can be reduced. Therefore, in this object position detection device, the amount of information to be stored can be reduced. Therefore, this object position detection apparatus can be introduced at a lower cost.

  An object position detection apparatus according to a tenth aspect of the present invention includes a swing imaging machine, an information storage unit, a captured image data receiving unit, a specific pixel coordinate deriving unit, a specific information table extracting unit, an actual floor polar deriving unit, and a 21st polar coordinate deriving unit. Prepare. The head swing imaging device has a lens unit and a turning drive mechanism. The turning drive mechanism can turn the lens unit in the yaw direction and the pitch direction. An information table is stored in the information storage unit. The information table is created for each combination of the yaw angle and pitch angle of the swing imaging device or for each pitch angle. In this information table, the actual floor polar coordinates are assigned to the pixel coordinates of the swing imaging device. The “actual floor polar coordinates” mentioned here are the polar coordinates of the actual floor surface. The photographed image data receiving unit receives photographed image data from the swing imaging device. The specific pixel coordinate deriving unit derives specific pixel coordinates. Here, the “specific pixel coordinates” are pixel coordinates of a specific part of the captured image data. The specific information table extracting means extracts the specific information table by collating the combination of the yaw angle and the pitch angle or the pitch angle with the information table. The “specific information table” here is a specific information table. The actual floor polar coordinate deriving means derives the actual floor polar coordinates using the specific pixel coordinates and the specific information table. The twenty-first polar coordinate deriving means derives the twenty-first polar coordinates using the actual floor polar coordinates derived by the actual floor polar coordinate deriving means.

  In this object position detection device, a swing imaging device is employed as a monitoring camera. For this reason, in this object position detection device, even when a large space room such as an office is to be monitored, there may be one or several monitoring cameras (when there is a shield in the office space or the like). That's enough. Therefore, this object position detection apparatus can be introduced at a lower cost than in the past. Further, in this object position detection device, the 21st polar coordinate deriving means derives the 21st polar coordinates using the actual floor polar coordinates derived by the actual floor polar coordinate deriving means. Therefore, in this object position detection device, for example, the polar coordinates of the individual feet of the monitoring subject can be derived from the polar coordinates of the head in the actual floor polar coordinates using the height of the individual of the monitoring subject. Therefore, in this object position detection apparatus, the presence position of the monitoring subject can be specified more accurately.

  An object position detection device according to an eleventh invention is the object position detection device according to the tenth invention, wherein the twenty-first polar coordinate deriving means is a predetermined distance from the distance component of the actual floor polar coordinate derived by the actual floor polar coordinate deriving means. Is added or subtracted to derive the 21st polar coordinate.

  In this object position detecting device, the 21st polar coordinate deriving means derives the 21st polar coordinate by adding or subtracting a predetermined distance from the distance component of the actual floor polar coordinate derived by the actual floor polar coordinate deriving means. For this reason, in this object position detection apparatus, the presence position of a monitoring subject can be specified simply and accurately.

  An object position detection apparatus according to a twelfth aspect includes a fisheye imager, an information storage unit, a captured image data receiving unit, a specific pixel coordinate deriving unit, an actual floor polar coordinate deriving unit, and a 21st polar coordinate deriving unit. An information table is stored in the information storage unit. In the information table, the actual floor polar coordinates are assigned to the pixel coordinates of the fish-eye imager. The “actual floor polar coordinates” mentioned here are the polar coordinates of the actual floor surface. The captured image data receiving unit receives captured image data from the fisheye imager. The specific pixel coordinate deriving unit derives specific pixel coordinates. Here, the “specific pixel coordinates” are pixel coordinates of a specific part of the captured image data. The actual floor polar coordinate deriving means collates the specific pixel coordinates with the information table to derive the actual floor polar coordinates. The twenty-first polar coordinate deriving means derives the twenty-first polar coordinates using the actual floor polar coordinates derived by the actual floor polar coordinate deriving means.

  In this object position detection device, a fish-eye imager is employed as a monitoring camera. For this reason, in this object position detection device, even when a large space room such as an office is to be monitored, there may be one or several monitoring cameras (when there is a shield in the office space or the like). That's enough. Therefore, this object position detection apparatus can be introduced at a lower cost than in the past. Further, in this object position detection device, the 21st polar coordinate deriving means derives the 21st polar coordinates using the actual floor polar coordinates derived by the actual floor polar coordinate deriving means. Therefore, in this object position detection device, for example, the polar coordinates of the individual feet of the monitoring subject can be derived from the polar coordinates of the head in the actual floor polar coordinates using the height of the individual of the monitoring subject. Therefore, in this object position detection apparatus, the presence position of the monitoring subject can be specified more accurately.

  In the object position detection device according to the first aspect of the present invention, even when a large space room such as an office is to be monitored, one or several monitoring cameras (when there is a shield in the office space or the like) If there is enough, it is enough. Therefore, this object position detection apparatus can be introduced at a lower cost than in the past.

  In the object position detection apparatus according to the second aspect of the invention, the first polar coordinates are quickly derived.

  In the object position detection apparatus according to the third aspect of the invention, the number of information tables that should be stored in the information storage unit can be reduced. Therefore, in this object position detection device, the amount of information to be stored can be reduced. Therefore, this object position detection apparatus can be introduced at a lower cost.

  In the object position detection device according to the fourth aspect of the invention, the number of polar coordinates assigned to pixel coordinates can be reduced. Therefore, in this object position detection device, the amount of information to be stored can be reduced. Therefore, this object position detection apparatus can be introduced at a lower cost.

  In the object position detection device according to the fifth aspect of the invention, the number of polar coordinates assigned to pixel coordinates can be reduced. Therefore, in this object position detection device, the amount of information to be stored can be reduced. Therefore, this object position detection apparatus can be introduced at a lower cost.

  In the object position detection apparatus according to the sixth aspect of the invention, the number of information tables that should be stored in the information storage unit can be reduced. Therefore, in this object position detection device, the amount of information to be stored can be reduced. Therefore, this object position detection apparatus can be introduced at a lower cost.

  In the object position detection device according to the seventh aspect of the present invention, even when a large space room such as an office is to be monitored, one or several monitoring cameras (when there is a shield in the office space or the like) If there is enough, it is enough. Therefore, this object position detection apparatus can be introduced at a lower cost than in the past.

  In the object position detection device according to the eighth aspect of the invention, the number of polar coordinates assigned to pixel coordinates can be reduced. Therefore, in this object position detection device, the amount of information to be stored can be reduced. Therefore, this object position detection apparatus can be introduced at a lower cost.

  In the object position detection device according to the ninth aspect of the invention, the number of polar coordinates assigned to pixel coordinates can be reduced. Therefore, in this object position detection device, the amount of information to be stored can be reduced. Therefore, this object position detection apparatus can be introduced at a lower cost.

  In the object position detection apparatus according to the tenth aspect of the invention, even when a large space room such as an office is to be monitored, one or several monitoring cameras (when there is a shield in the office space or the like) If there is enough, it is enough. Therefore, this object position detection apparatus can be introduced at a lower cost than in the past. Further, in this object position detection device, the 21st polar coordinate deriving means derives the 21st polar coordinates using the actual floor polar coordinates derived by the actual floor polar coordinate deriving means. Therefore, in this object position detection device, for example, the polar coordinates of the individual feet of the monitoring subject can be derived from the polar coordinates of the head in the actual floor polar coordinates using the height of the individual of the monitoring subject. Therefore, in this object position detection apparatus, the presence position of the monitoring subject can be specified more accurately.

  With the object position detection device according to the eleventh aspect of the present invention, it is possible to specify the presence position of the monitoring subject simply and accurately.

  In the object position detection apparatus according to the twelfth aspect of the present invention, even when a large space room such as an office is to be monitored, one or several monitoring cameras (when there is an obstacle in the office space or the like) If there is enough, it is enough. Therefore, this object position detection apparatus can be introduced at a lower cost than in the past. Further, in this object position detection device, the 21st polar coordinate deriving means derives the 21st polar coordinates using the actual floor polar coordinates derived by the actual floor polar coordinate deriving means. Therefore, in this object position detection device, for example, the polar coordinates of the individual feet of the monitoring subject can be derived from the polar coordinates of the head in the actual floor polar coordinates using the height of the individual of the monitoring subject. Therefore, in this object position detection apparatus, the presence position of the monitoring subject can be specified more accurately.

  Here, the human position information deriving device 1 according to the embodiment of the present invention will be described with reference to the drawings of FIGS.

  A person position information deriving device 1 according to an embodiment of the present invention is mainly composed of a head monitoring camera 2 and an information processing device 3, as shown in FIG. Hereinafter, each of these components will be described in detail.

(1) Swing monitoring camera As shown in FIG. 2, the swing monitoring camera 2 according to the present embodiment is a monocular monitoring camera, and mainly includes a body 21, a lens unit 22, and a drive mechanism unit 23. It is composed of The body 21 accommodates the lens unit 22 and the drive mechanism unit 23. The lens unit 22 is attached to the drive mechanism unit 23. The lens unit 22 can be rotated by a driving mechanism unit 23 with a predetermined angle step about a first axis X1 that is an axis substantially parallel to the vertical direction when the body 21 is attached to the ceiling. Yes. Further, the lens unit 22 is driven by the drive mechanism unit 23 along a semicircular orbit in predetermined angular increments about the first axis X1 and the second axis X2, which is an axis orthogonal to the optical axis of the lens unit 22. It is movable. Hereinafter, the rotational movement angle of the lens unit 22 around the first axis X1 is referred to as a yaw angle, and the rotational movement angle of the lens unit 22 around the second axis X2 is referred to as a pitch angle. In addition, a camera control unit (not shown) is provided inside the swing monitoring camera 2. The lens unit 22 is set so as to periodically move along a prescribed route by the camera control unit. The camera control unit transmits the image data to the information processing apparatus 3 every time an image is taken, and at this time, the camera control unit is set to transmit the yaw angle and pitch angle information of the lens unit 22 at that time. Has been.

(2) Information processing apparatus As shown in FIG. 3, the information processing apparatus 3 according to the present embodiment mainly includes a central processing unit 31, a main memory 33, a hard disk 34, a connection unit 32, an IDE interface 35, and a LAN. The interface 36 is configured. When maintenance or update is required for the information processing device 3, an input device (not shown), a display (not shown), or the like is appropriately connected to an input device interface or a display interface (not shown). In the information processing apparatus 3, the central processing unit 31 passes the first bus line 37, the main memory 33 passes the second bus line 38, and the LAN interface 36 and the IDE interface 35 pass the third bus line 39. To the connection part 32. Here, the central processing unit 31 is, for example, a semiconductor chip called a microprocessor, and mainly includes a control unit 31A and a calculation unit 31B (in addition to a primary cache memory, a secondary cache memory, and the like). May be included). The main memory 33 is, for example, a semiconductor chip such as a RAM (Random Access Memory). The connection unit 32 is a semiconductor chip such as a chip set. The hard disk 34 may be an external type. As shown in FIG. 4, the hard disk 34 stores programs such as an operating system 34a, a device driver 34b, and a human position information deriving application 34c. The operating system 34a is, for example, WINDOWS (registered trademark), MAC OS (registered trademark), OS / 2, UNIX (registered trademark) (for example, Linux (registered trademark)) or BeOS (registered trademark). It performs hardware management of the units 32 to 34 and various interfaces 35 and 36, provision of a user interface, management of various data, processing of common parts of applications, and the like. The device driver 34 b is a dedicated program prepared for each of the hard disk 34 and the connection unit 32, and performs a bridge for the operating system 34 a to control the hard disk 34 and the connection unit 32. The person position information deriving application 34c is a program for detecting the position of a person from the image data obtained from the head monitoring camera 2, and mainly includes a head coordinate deriving module 34d and a polar coordinate conversion module 34e. Yes. The person position information deriving application 34c stores a polar coordinate conversion table Tm1 as shown in FIG. In the present embodiment, the polar coordinate conversion table Tm1 is created for each pitch angle of the swing monitoring camera 2. If the pitch angle of the head surveillance camera 2 is the same, it is not necessary to change the allocation of polar coordinates to pixel coordinates even if the yaw angle changes, but if the pitch angle changes even if the yaw angle is the same. This is because the assignment of polar coordinates to pixel coordinates must be changed greatly. In the polar coordinate conversion table Tm1, as shown in FIG. 5, the pixel coordinates of the head monitoring camera 2 are associated with polar coordinates. In the present embodiment, the association between the pixel coordinates and the polar coordinates is determined in consideration of the actual polar coordinates of the floor surface and the average height of a person. This association method will be described in detail later. In the present embodiment, the origin of the pixel coordinates is defined at the center pixel of the swing monitoring camera 2, and the value of the yaw direction component of the pixel coordinates is 0 and a positive value as shown in FIG. The polar coordinates are associated only with the pixel coordinates (hatched pixel coordinates) EC1. At this time, polar coordinates having an angle component of 0 ° are assigned to center pixel coordinates having a pitch direction component value of 0. That is, in the present embodiment, polar coordinates having an angle component of 0 and a positive value are associated with pixel coordinates. An IDE (Integrated Drive Electronics) interface 35 connects the hard disk 34 to the connection unit 32. The LAN interface 36 is communicatively connected to the head surveillance camera 2 via the Ethernet (registered trademark) cable 4 (see FIGS. 1 and 2).

  Next, the operation of the information processing apparatus 3 will be described with reference to FIG.

  As shown in FIG. 7, the control unit 31A reads a program temporarily stored in the main memory 33 (see Fd5), and instructs the units 32 to 34 according to the read program (see Fc1 to Fc3). The calculation unit 31B acquires necessary data from the main memory 33 according to the instruction of the control unit 31A (see Fd1) and performs calculation processing (for example, arithmetic calculation processing, logical calculation processing, etc.). The main memory 33 acquires programs and data from the hard disk 34 (see Fd3) and temporarily stores them, and temporarily stores data (see Fd2 and Fd7) transmitted from the calculation unit 31B and the swing monitoring camera 2. To do. In addition, the main memory 33 transmits data or the like temporarily stored in accordance with an instruction from the control unit 31A to the units 32 to 34 and the swing monitoring camera 2 (see Fd1, Fd4, and Fd6). The hard disk 34 supplies a program, data, and the like to the main memory 33 (see Fd3) and stores data transmitted from the main memory 33 (see Fd4) in accordance with an instruction from the control unit 31A.

<Method for associating pixel coordinates and polar coordinates of the head surveillance camera>
In the present embodiment, the association between the pixel coordinates and the polar coordinates of the head monitoring camera is performed as follows.

  First, a polar coordinate mat 50 (see FIG. 8) larger than the size of the room RM to be monitored is produced. Next, a plate member capable of supporting the polar coordinate mat 50 without being curved is laid in the room RM. Subsequently, the polar mat 50 is brought into the room to be monitored, and the polar mat 50 is spread on the plate member. At this time, the polar coordinate mat 50 is installed so that the center point of the polar coordinates (the point at which the distance component is 0) is directly below the swing monitoring camera 2. Subsequently, the polar mat 50 supported by the plate member is lifted to the average height. Then, in this state, the head monitoring camera 2 is operated, and images are taken at different pitch angles. At this time, imaging is performed so that one of the radiation in the polar mat 50 passes through the center of the captured image. Thereafter, edge extraction image processing is executed in the information processing apparatus 3, and pixel coordinates corresponding to the intersection portion of the polar coordinate mat 50 are derived. Then, polar coordinates of the intersection are assigned to the pixel coordinates. In addition, since the step of the polar coordinate scale is determined when the polar coordinate mat 50 is manufactured, the intersection coordinates on the polar coordinate mat 50 are uniquely determined. After intersections of all polar coordinate mats 50 are assigned to pixel coordinates, polar coordinates are assigned to the remaining pixel coordinates based on the pixel coordinates to which polar coordinates are assigned. At this time, this allocation may be performed manually or automatically using computer technology.

<People location information derivation process>
In the human position information deriving device 1 according to the embodiment of the present invention, absolute value conversion processing is executed according to the flowchart shown in FIG. 9, polar coordinate conversion table specifying processing is executed according to the flowchart shown in FIG. The human position information derivation process is executed according to the flowchart shown. In the human position information deriving device 1, these processes are executed every time image data, yaw angle data, and pitch angle data are transmitted from the head monitoring camera 2.

  In FIG. 9, in step S <b> 1, when the information processing device 3 receives image data from the head monitoring camera 2, the information processing device 3 derives position data (hereinafter referred to as “head position data”) of the black portion on the image data. The process of step S1 is performed by conventional head detection technology (for example, Japanese Patent Application Laid-Open No. 2003-189295, Dana H. Ballard, Christopher M. Brown, Yasuo Fukumura [etc.], “Computer Vision”, Japan. (See Computer Association etc.) Specifically, a method of performing edge extraction processing after binarization of image data and detecting the head by Hough transform, or HIS conversion of RGB in the head region after detecting the head by frame difference and a Hue value (hue) And a method using the center of gravity of the maximum value of the head as a head coordinate (a technology of Sony Corporation).

  In step S2, the information processing device 3 converts the yaw direction component of the head position data into an absolute value. Hereinafter, head position data in which the yaw direction component is converted into an absolute value is referred to as converted head position data.

  In step S <b> 3, the information processing device 3 temporarily stores the head position data and the converted head position data in the main memory 33.

  In FIG. 10, in step S11, when the information processing device 3 receives the pitch angle data from the swing monitoring camera 2, from the plurality of polar coordinate conversion tables Tm1, the polar coordinate conversion corresponding to the pitch angle data received in step S11. The table Tm1 is extracted.

  In step S12, the information processing apparatus 3 temporarily stores the polar coordinate conversion table Tm1 extracted in step S11 in the main memory 33.

  In FIG. 11, in step S21, the information processing device 3 reads the converted head position data temporarily stored in the main memory 33 in step S3, and stores the converted head position data in the main memory 33 in step S12. The data is converted into polar coordinate data by collating with the temporarily stored polar coordinate conversion table Tm1.

  In step S22, the information processing apparatus 3 reads the head position data temporarily stored in the main memory 33 in step S3, and determines whether or not the yaw direction component of the head position data is a negative value. As a result of the determination by the information processing device 3 in step S22, when the yaw direction component of the head position data is a negative value, the process proceeds to step S23. On the other hand, when the yaw direction component of the head position data is not a negative value as a result of the determination by the information processing device 3 in step S22, that is, when the yaw direction component of the head position data is 0 or a positive value, The processing moves to step S24.

  In step S23, the information processing apparatus 3 performs code conversion on the angle component of the polar coordinate data derived in step S21.

  In step S24, the information processing apparatus 3 adds the yaw angle sent from the swing monitoring camera 2 to the polar coordinate data derived in step S21 or the angle component of the polar coordinate data derived in step S23.

  In step S25, the information processing apparatus 3 stores the polar coordinate data derived in step S24 in the hard disk 34.

<Conversion processing from polar coordinates to Cartesian coordinates>
The polar coordinate data (r, θ) derived in step S25 can be converted into orthogonal coordinate data (x, y) by the following equation (1). Here, r is the distance from the camera-corresponding position on the floor, and θ is an angle centered on the camera-corresponding position on the floor.

  (X, y) = (r cos θ, r sin θ) (1)

<Visualization of human location data>
As described above, if the image data is synthesized based on the polar coordinates and the human position data is plotted on the image data, a distribution map of people in the room can be obtained.

<Features of human position information deriving device>
(1)
In the human position information deriving device 1 according to the present embodiment, a head surveillance camera 2 is employed as the surveillance camera. For this reason, in this person position information deriving device 1, even when a large space room such as an office is to be monitored, one or several monitoring cameras (when there is an obstacle in the office space or the like) If there is, it is enough. Therefore, this person position information deriving device 1 can be introduced at a lower cost than in the past.

(2)
In the human position information deriving device 1 according to the present embodiment, a monocular swing monitoring camera 2 is employed. For this reason, in this human position information deriving device 1, the occlusion phenomenon (a phenomenon in which a portion that can be seen and a portion that cannot be seen depending on the position of the surveillance camera) that is a problem in a stereo surveillance camera or a multi-view surveillance camera does not matter.

(3)
In the human position information deriving device 1 according to the present embodiment, in the polar coordinate conversion table Tm1, the pixel coordinates on the plus side in the yaw direction out of the pixel coordinates of the head monitoring camera 2 (0 when the center in the yaw direction is 0). Polar coordinates are assigned only to (including). In the human position information deriving device 1, the information processing device 3 converts the yaw direction component of the head position data into an absolute value and collates the head position data obtained by converting the yaw direction component into an absolute value against the polar coordinate conversion table Tm1. To polar coordinate data. In the human position information deriving device 1, the information processing device 3 further determines whether or not the yaw direction component of the head position data is a negative value, and the yaw direction component of the head position data is a negative value. If it is, the angle component of the polar coordinate data derived in step S21 is sign-converted. That is, in the human position information deriving device 1, the number of polar coordinates assigned to the pixel coordinates is reduced. For this reason, in this person position information deriving device 1, the amount of information to be stored is reduced. Therefore, this person position information deriving device 1 can be introduced at low cost.

(4)
In the human position information deriving device 1 according to the present embodiment, a polar coordinate conversion table Tm1 is created for each pitch angle. In the human position information deriving device 1, when the information processing device 3 receives the pitch angle data from the head monitoring camera 2, the polar coordinates corresponding to the pitch angle data received in step S11 from the plurality of polar coordinate conversion tables Tm1. The conversion table Tm1 is extracted. In the human position information deriving device 1, the information processing device 3 transmits the yaw sent from the head monitoring camera 2 to the polar coordinate data derived in step S21 or the angle component of the polar coordinate data derived in step S23. Add corners. That is, in the human position information deriving device 1, the number of polar coordinate conversion tables Tm1 that should be stored in the hard disk 34 is reduced. For this reason, in this person position information deriving device 1, the amount of information to be stored is reduced. Therefore, this person position information deriving device 1 can be introduced at a cost.

<Modification>
(A)
In the human position information deriving device 1 according to the previous embodiment, the single-lens swing monitoring camera 2 is employed, but a compound-eye swing monitoring camera 102 as shown in FIG. 12 may be employed as the surveillance camera. Absent.

(B)
In the human position information deriving device 1 according to the previous embodiment, the monocular swing monitoring camera 2 is employed, but a fish-eye surveillance camera may be employed as the surveillance camera. In such a case, the entire floor of the living room is imaged at once. For this reason, only one polar coordinate conversion table is required. In such a case, polar coordinates may be associated only with pixel coordinates in which the horizontal component value is 0 or a positive value in the pixel coordinates, and in the pixel coordinates, the vertical component value is 0 or a positive value. Polar coordinates may be associated only with the pixel coordinates, or polar coordinates may be associated with all pixel coordinates.

(C)
In the human position information deriving device 1 according to the previous embodiment, the yaw direction component of the central pixel coordinate is set to 0, but when the number of pixels in the yaw direction of the head monitoring camera 2 is an even number, FIG. The yaw direction component of the center pixel coordinate may be set to 1 and −1 as shown in FIG.

(D)
In the human position information deriving device 1 according to the previous embodiment, only the pixel coordinates (hatched pixel coordinates) EC1 in which the value of the yaw direction component is 0 and positive among the pixel coordinates of the head monitoring camera 2 are polar coordinates. Are associated with each other, but polar coordinates may be associated with all pixel coordinates. In such a case, the amount of information to be stored increases.

(E)
In the human position information deriving device 1 according to the previous embodiment, the origin of the pixel coordinates is defined for the central pixel of the swing monitoring camera 2. However, the origin of the pixel coordinates is the number of yaw direction components having a positive value. It may be freely defined as long as the restriction that it exceeds the number of yaw direction components taking a negative value is observed.

(F)
In the human position information deriving device 1 according to the previous embodiment, the absolute value processing is executed. However, the absolute value processing is eliminated and the human position information deriving processing is executed according to the flowchart shown in FIG. Good. In such a case, the information processing device 3 derives head position data in advance and temporarily stores the head position data in the main memory 33.

  In FIG. 14, in step S31, the information processing apparatus 3 includes pixel coordinate data that matches the head position data temporarily stored in the main memory 33 in the polar coordinate conversion table Tm1 temporarily stored in the main memory 33 in step S12. It is determined whether or not exists. As a result of the determination by the information processing apparatus 3 in step S31, when pixel coordinate data that matches the head position data exists in the polar coordinate conversion table Tm1, the process proceeds to step S32. As a result of the determination by the information processing device 3 in step S31, when there is no pixel coordinate data that matches the head position data in the polar coordinate conversion table Tm1, the process proceeds to step S33.

  In step S32, the information processing apparatus 3 collates the head position data with the polar coordinate conversion table Tm1 and converts it into polar coordinate data.

  In step S33, the information processing device 3 performs code conversion on the yaw direction component of the head position data. Hereinafter, head position data obtained by sign-converting the yaw direction component is referred to as converted head position data.

  In step S34, the information processing apparatus 3 collates the converted head position data with the polar coordinate conversion table Tm1 and converts it into polar coordinate data.

  In step S35, the information processing device 3 performs code conversion on the angle component of the polar coordinate data derived in step S34.

  In step S36, the information processing device 3 adds the yaw angle sent from the head monitoring camera 2 to the polar coordinate data derived in step S32 or the angle component of the polar coordinate data derived in step S35.

  In step S37, the information processing device 3 stores the polar coordinate data derived in step S36 in the hard disk 34.

(G)
In the human position information deriving device 1 according to the previous embodiment, the polar coordinate conversion table Tm1 is created for each pitch angle of the swing monitoring camera 2, but a polar coordinate conversion table exists for each combination of yaw angle and pitch angle. It may be. In such a case, the amount of information to be stored increases.

(H)
Although not particularly mentioned in the previous embodiment, when a wide-angle lens is employed for the lens unit 22, it is preferable to correct distortion. The distortion aberration D is defined by the following equation (2).

D = ((Rr−Ri) / Ri) × 100 (2)
Here, Ri is an ideal image height, and Rr is an actual image height (see FIG. 15).

  Therefore, in order to calculate the coordinate value corresponding to the pixel coordinate, first, the pixel coordinate (Ur, Vr) is converted into the pixel coordinate (Ui, Vi) before being distorted, and the perspective transformation is performed from the camera position. You need to find the coordinates.

  The pixel coordinates without distortion are defined by the following equation (3).

(Ui, Vi) = (Ur / (1 + D), Vr / (1 + D)) (3)
Here, D = A0 + A1 · Y ¹ + A2 · Y ² +... An · Y ⁿ .

  Further, the pixel coordinates without distortion are also defined by the following equation (4).

(Ui, Vi) = (d k · {1 + κ · ((Ur-Uc) 2 + (Vr-Vc) 2} · (Ur-Uc), d k · {1 + κ · ((Ur-Uc) 2 + ( Vr−Vc) ² } · (Vr−Vc)) (4)
Here, d _k is a distance, Uc and Vc are center pixel coordinates, and κ is a distortion aberration coefficient.

(I)
In the human position information deriving device 1 according to the previous embodiment, the polar mat 50 is picked up with the polar mat 50 supported by the plate member raised to the average height, and the image data is subjected to edge extraction processing. The pixel coordinates and polar coordinates of the swing monitoring camera are associated with each other, but the position of the swing monitoring camera and the yaw angle and pitch angle of the swing monitoring camera are calculated from the data such as the yaw angle and pitch angle of the swing monitoring camera. The association between pixel coordinates and polar coordinates may be performed. Note that a specific calculation method can be derived without undue effort by those skilled in the art, and thus description thereof is omitted here.

(J)
Although not particularly mentioned in the previous embodiment, pixel coordinates when the optical axis of the lens unit 22 is parallel to the vertical direction, that is, when an image of the floor surface directly below the head monitoring camera 2 is captured. May be assigned Cartesian coordinates or special coordinates as shown in FIGS. 16, 17 and 18 instead of polar coordinates. Note that the special coordinate feature of FIG. 16 is that the central portion is a single circular grid, and the special coordinate feature of FIG. 17 is a coordinate in which the central portion is an orthogonal coordinate and the surrounding portion takes distortion into consideration. The special coordinates in FIG. 18 are characterized in that the central portion is orthogonal coordinates and the surrounding portions are polar coordinates.

(K)
In the human position information deriving device 1 according to the previous embodiment, the average height of the person is taken into consideration and the pixel coordinates and polar coordinates of the head surveillance camera are associated with each other, and the position of the person is detected based on the association. However, the pixel coordinates and polar coordinates of the swing monitoring camera may be associated with the average height of the person sitting, and the position of the person may be detected based on the association. . In such a case, “position detection is performed based on the association that takes into account the average height of the person sitting when the person's position data belongs within a certain range centered on the coordinate data of the chair for a certain period of time. In other cases, position detection is performed based on association that takes into account the average height of a person ”. This is because, in an office or the like, when a person stays in the vicinity of a chair for a certain period of time, the person is usually seated.

  Further, it may be determined by image analysis technology from image data whether a person is in a sitting state or a standing state, and which association is used in accordance with the state.

(L)
In the human position information deriving device 1 according to the previous embodiment, the average height of the person is taken into consideration and the pixel coordinates and polar coordinates of the head surveillance camera are associated with each other, and the position of the person is detected based on the association. However, the position of the person may be corrected after taking into account the height of the individual. In such a case, the distance component of the polar coordinate data derived in step S24 is corrected with a value that takes into account the individual's height. Individual height data can be specified by preparing a table or the like in which individual IDs and height data are associated with each other. Specifically, it is conceivable to derive a personal ID by using a wireless LAN technique or an image analysis technique, and to derive height data by comparing the personal ID with a previous table.

  As a specific example, it is conceivable to use a human position detection technique using a wireless LAN. A person carries a wireless LAN terminal, such as a mobile phone with a wireless LAN function, and always sends a personal ID from the mobile phone. Then, the position of the person is detected and at the same time the personal ID data of the person is specified. When the position data output from the human position information deriving device according to the present invention is compared with the position data output from the human position detecting device using the wireless LAN, and the position data is within a certain distance range. Corrects the position data output from the human position information deriving device according to the present invention with the height data obtained by comparing the personal ID data with the table. In this way, more precise position detection can be performed.

  In addition, if a table or the like in which the personal ID is associated with the height data at the time of sitting or a seating determination algorithm as shown in the modification (K) is prepared, the person is seated. Even so, it is possible to detect the exact position of the person.

(M)
In the human position information deriving device 1 according to the previous embodiment, the average height of the person is taken into consideration and the pixel coordinates and polar coordinates of the head surveillance camera are associated with each other, and the position of the person is detected based on the association. However, the pixel coordinates of the head monitoring camera may be associated with the actual polar coordinates on the floor, and then the position of the person may be detected by taking into account the individual's height. In such a case, the polar coordinate mat 50 supported by the plate member does not need to be lifted to the average height, and the swing monitoring camera 2 is operated with the polar coordinate mat 50 laid on the floor surface, and imaging is performed at different pitch angles. Is done. Then, the distance component of the polar coordinate data derived in step S24 is corrected with a value that takes into account the individual's height. Individual height data can be specified by preparing a table or the like in which individual IDs and height data are associated with each other. Specifically, it is conceivable to derive a personal ID by using a wireless LAN technique or an image analysis technique, and to derive height data by comparing the personal ID with a previous table.

  The object position detection apparatus according to the present invention may include one or several monitoring cameras (when there is an obstacle in the office space or the like) even when a large space room such as an office is to be monitored. That's enough. Therefore, this object position detection device has a feature that it can be introduced at a lower cost than conventional ones, and is useful as an object position detection device for a relatively large space such as an office. Further, such an object position detection device is expected to be applied to air conditioning in a large space room such as an office.

It is a schematic block diagram of the person location information derivation device concerning an embodiment of the invention. 1 is an external perspective view of a head surveillance camera employed in a human position information deriving device according to an embodiment of the present invention. It is a block diagram of the information processing apparatus employ | adopted for the person location information derivation | leading-out apparatus which concerns on embodiment of this invention. It is an image figure of the hard disk of the information processing apparatus employ | adopted for the person location information derivation | leading-out apparatus which concerns on embodiment of this invention. It is an image figure of the polar coordinate conversion table memorize | stored in the hard disk of the information processing apparatus employ | adopted as the person position information derivation | leading-out apparatus which concerns on embodiment of this invention. It is an image figure of the pixel coordinate of the swing monitoring camera employ | adopted for the person position information derivation | leading-out apparatus which concerns on embodiment of this invention. It is a figure showing the flow of the control signal and data in the person position information derivation device concerning an embodiment of the invention. It is a top view of the polar coordinate mat used for matching with the pixel coordinate of a head surveillance camera, and a polar coordinate in the person position information derivation device concerning an embodiment of the invention. It is a flowchart showing the flow of the absolute value process performed in the person location information derivation device concerning an embodiment of the invention. It is a flowchart showing the flow of the polar coordinate conversion table specific process performed in the person position information derivation device concerning an embodiment of the invention. It is a flowchart showing the flow of the person position information derivation | leading-out process performed in the person position information derivation | leading-out apparatus which concerns on embodiment of this invention. It is an external appearance perspective view of the compound eye swing monitoring camera which concerns on a modification (A). It is an image figure of the pixel coordinate of the head monitoring camera employ | adopted as the person position information derivation | leading-out apparatus which concerns on a modification (C). It is a flowchart showing the flow of the person position information derivation | leading-out process performed in the person position information derivation | leading-out apparatus concerning a modification (E). It is a reference figure which shows the principle of the distortion aberration correction process performed in the person position information derivation | leading-out apparatus which concerns on a modification (G). It is an image figure of the special coordinates employ | adopted in the person position information derivation device concerning modification (J). It is an image figure of a special coordinate in the person position information derivation device concerning modification (J). It is an image figure of a special coordinate in the person position information derivation device concerning modification (J).

Explanation of symbols

1 Person position information deriving device (object position detecting device)
2 Swing monitoring camera (swing camera)
22 Lens part 23 Drive mechanism part (turning drive mechanism)
34 Hard disk (information storage unit)
34d Head coordinate deriving module (specific pixel coordinate deriving means)
34e Polar coordinate conversion module (first polar coordinate deriving means)
36 LAN interface (captured image data receiver)
Tm1 Polar coordinate conversion table (information table)

Claims (12)

A swing imaging device (2) having a lens unit (22) and a turning drive mechanism (23) capable of turning the lens unit in a yaw direction and a pitch direction;
An information storage unit (Tm1) that stores an information table (Tm1) that is created for each combination of yaw angle and pitch angle of the lens unit or for each pitch angle and that assigns first polar coordinates to pixel coordinates of the swing imaging device. 34)
A captured image data receiving unit (36) for receiving captured image data from the swing imaging device;
Specific pixel coordinate deriving means (34d) for deriving pixel coordinates (hereinafter referred to as "specific pixel coordinates") of a specific portion of the captured image data;
Specific information table extraction means for extracting the specific information table (hereinafter referred to as “specific information table”) by comparing the yaw angle and pitch angle or the pitch angle with the information storage unit;
An object position detecting device (1) comprising: first polar coordinate deriving means (34e) for deriving the first polar coordinate using the specific pixel coordinates and the specific information table. The object position detection device according to claim 1, wherein the first polar coordinate deriving unit derives the first polar coordinate by collating the specific pixel coordinate with the specific information table. The information table is created for each pitch angle, and assigns the first polar coordinates to the pixel coordinates of the swing imaging device,
The specific information table extracting means extracts the specific information table by collating the pitch angle with the information storage unit,
The object according to claim 1, further comprising second polar coordinate deriving means for deriving a second polar coordinate by adding or subtracting the yaw angle to an angle component of the first polar coordinate derived by the first polar coordinate deriving means. Position detection device. The pixel coordinates are orthogonal coordinates in which the central part in the yaw direction is 0, or 1 and −1,
The information table is created for each combination of yaw angle and pitch angle of the lens unit or for each pitch angle, and is half of the yaw direction (the center part of the yaw direction is the pixel coordinate of the swing imaging device). 1st polar coordinates are assigned only to pixel coordinates (hereinafter referred to as “half pixel coordinates”) of 0 (including 0 if 0),
When the specific pixel coordinate does not coincide with any pixel coordinate of the half-pixel coordinate, further comprising first code conversion means for code-converting the yaw direction component of the specific pixel coordinate,
The first polar coordinate deriving unit specifies the specific pixel coordinate whose yaw direction component is code-converted by the first code converting unit when the specific pixel coordinate does not match any pixel coordinate of the half-pixel coordinate. Deriving the first polar coordinates by collating with an information table;
Second code conversion means for code-converting the angle component of the first polar coordinate derived by the first polar coordinate deriving means when the specific pixel coordinate does not match any of the half-pixel coordinates; The object position detection device according to claim 1, further comprising: The pixel coordinates are orthogonal coordinates in which the central part in the yaw direction is 0, or 1 and −1,
The information table is created for each combination of the yaw angle and pitch angle of the lens unit or for each pitch angle, and the pixel coordinates on the plus side of the yaw direction (yaw direction) among the pixel coordinates of the swing imaging device The first polar coordinate is assigned only to (including 0 if the center of is 0),
Absolute value converting means for converting the yaw direction component of the specific pixel coordinates into an absolute value;
A first polar coordinate deriving unit for deriving the first polar coordinate by collating a specific pixel coordinate whose yaw direction component has been converted into an absolute value by the absolute value converting unit with the specific information table;
A yaw direction component code determination unit that determines whether or not the yaw direction component of the specific pixel coordinate is a negative value, and the yaw direction component code determination unit determines that the yaw direction component of the specific pixel coordinate is a negative value. 3. The object position detection device according to claim 1, further comprising: a third code conversion unit configured to code-convert an angle component of the first polar coordinate derived by the first polar coordinate deriving unit. The information table is created for each pitch angle,
The specific information table extracting means extracts the specific information table by collating the pitch angle with the information storage unit,
The system further comprises third polar coordinate deriving means for deriving a third polar coordinate by adding or subtracting the yaw angle to or from the angular component of the first polar coordinate code converted by the second code converting means or the third code converting means. Item 6. The object position detection device according to Item 4 or 5. A fisheye imager;
An information storage unit for storing an information table for assigning eleventh polar coordinates to pixel coordinates of the fisheye imager;
A captured image data receiving unit for receiving captured image data from the fisheye imager;
Specific pixel coordinate deriving means for deriving pixel coordinates of a specific part of the captured image data (hereinafter referred to as “specific pixel coordinates”);
An object position detecting device comprising: eleventh polar coordinate deriving means for deriving the eleventh polar coordinate by comparing the specific pixel coordinate with the information table. The pixel coordinates are orthogonal coordinates in which the central portion in the horizontal direction or the vertical direction is 0, 1 or −1,
The information table includes pixel coordinates (hereinafter, “half pixel coordinates”) of pixel coordinates of the fish-eye imager, including half in the horizontal direction or the vertical direction (including 0 when the central portion in the horizontal direction or the vertical direction is 0). The eleventh polar coordinate is assigned only to
An eleventh code conversion means for code-converting a horizontal component or a vertical component of the specific pixel coordinate when the specific pixel coordinate does not match any pixel coordinate of the half-pixel coordinate;
The eleventh polar coordinate deriving means, when the specific pixel coordinate does not coincide with any one of the half-pixel coordinates, the specific pixel coordinate whose horizontal component or vertical component is code-converted by the eleventh code conversion means. To the information table to derive the eleventh polar coordinate,
Twelfth code conversion means for code-converting the angle component of the first polar coordinate derived by the eleventh polar coordinate deriving means when the specific pixel coordinate does not coincide with any one of the half-pixel coordinates. The object position detection device according to claim 7. The pixel coordinates are orthogonal coordinates in which the central portion in the horizontal direction or the vertical direction is 0, 1 or −1,
The information table includes eleventh polar coordinates only in the pixel coordinates on the plus side in the horizontal direction or the vertical direction among pixel coordinates of the fisheye imager (including 0 when the central portion in the horizontal direction or the vertical direction is 0). allocation,
An absolute value converting means for converting the horizontal component or the vertical component of the specific pixel coordinate into an absolute value;
The eleventh polar coordinate deriving means derives the eleventh polar coordinate by collating the specific pixel coordinates whose horizontal component or vertical component is absolute valued by the absolute value converting means with the information table,
The horizontal component or the like sign determining means for determining whether the horizontal component or the vertical component of the specific pixel coordinate is a negative value, and the horizontal or vertical direction of the specific pixel coordinate by the horizontal component equal sign determining means is a negative value. The object position detection apparatus according to claim 7, further comprising: thirteenth code conversion means for code-converting the angle component of the eleventh polar coordinate derived by the eleventh polar coordinate deriving means when it is determined that A swing imaging device having a lens part and a turning drive mechanism capable of turning the lens part in a yaw direction and a pitch direction;
It is created for each combination of the yaw angle and pitch angle of the swing imaging device or for each pitch angle, and the polar coordinates of the actual floor surface (hereinafter referred to as “actual floor polar coordinates”) as the pixel coordinates of the swing imaging device An information storage unit for storing an information table to which
A photographed image data receiving unit for receiving photographed image data from the head image pickup device;
Specific pixel coordinate deriving means for deriving pixel coordinates of a specific part of the captured image data (hereinafter referred to as “specific pixel coordinates”);
A specific information table extracting means for extracting the specific information table (hereinafter referred to as “specific information table”) by comparing the combination of the yaw angle and the pitch angle or the pitch angle with the information table;
An actual floor polar coordinate deriving unit for deriving the actual floor polar coordinate using the specific pixel coordinate and the specific information table, and a 21st polar coordinate is derived using the actual floor polar coordinate derived by the actual floor polar coordinate deriving unit. An object position detecting device comprising: 21st polar coordinate deriving means. The object position according to claim 10, wherein the twenty-first polar coordinate deriving unit derives the twenty-first polar coordinate by adding or subtracting a predetermined distance from a distance component of the actual floor polar coordinate derived by the actual floor polar coordinate deriving unit. Detection device. A fisheye imager;
An information storage unit for storing an information table for assigning actual floor polar coordinates (hereinafter referred to as “actual floor polar coordinates”) to the pixel coordinates of the fish-eye imager;
A captured image data receiving unit for receiving captured image data from the fisheye imager;
Specific pixel coordinate deriving means for deriving pixel coordinates of a specific part of the captured image data (hereinafter referred to as “specific pixel coordinates”);
Real floor polar coordinate deriving means for deriving the actual floor polar coordinates by collating the specific pixel coordinates with the information table;
An object position detecting device comprising: 21st polar coordinate deriving means for deriving 21st polar coordinates using the actual floor polar coordinates derived by the actual floor polar coordinate deriving means.

Images