WO2011155439A1

WO2011155439A1 - Motion detection device

Info

Publication number: WO2011155439A1
Application number: PCT/JP2011/062927
Authority: WO
Inventors: 圭介淺利; 本郷　仁志
Original assignee: 三洋電機株式会社
Priority date: 2010-06-07
Filing date: 2011-06-06
Publication date: 2011-12-15
Also published as: JP2011257862A

Abstract

A CPU (14p) specifies a plurality of partitioned areas upon a map image representing a bird's eye view state of a plane defining a space, and assigns a plurality of measurement areas which respectively correspond to the specified plurality of partitioned areas to a field of view image which has been output from a camera (12) which captures the plane. The CPU (14p) also classifies the measurement areas which have been assigned to the camera image into proper areas which match size conditions and non-proper areas which deviate from the size conditions. Degrees of congestion of partial fields of view belonging to the proper areas are detected on the basis of the field of view image which has been output from the camera (12), whereas degrees of congestion of partial fields of view belonging to the non-proper areas are detected on the basis of the degrees of congestion that have been detected for the proper areas. Loss of motion detection precision caused by a decrease in sizes of the measurement areas is suppressed.

Description

Motion detection device

The present invention relates to a motion detection device, and more particularly to a motion detection device that detects the motion of one or more objects existing on a plane based on an object scene image (or subject image) output from a camera that captures the plane. .

An example of this type of device is disclosed in Patent Document 1. According to this background art, the camera captures a plane in which a person walks from an oblique direction. An image output from the camera is divided into a plurality of motion processing areas. At this time, the size of the motion processing area allocated to the front scene (the front part of the three-dimensional space captured by the camera) is set to the rear scene (the far part of the three-dimensional space captured by the camera). It becomes larger than the size of the motion processing area to be allocated. The motion of the image is detected for each motion processing area and compared with a threshold value set for each motion processing area. The degree of congestion of the walking person is estimated based on the comparison result thus obtained.

JP 2009-110152 A

However, in the background art, it may be difficult to detect motion depending on the set size of the motion processing area.

Therefore, a main object of the present invention is to provide a motion detection device that can suppress a decrease in motion detection accuracy.

The motion detection apparatus according to the present invention (10: reference numeral corresponding to the embodiment; the same applies hereinafter) designates a plurality of first areas on a reference image representing a bird's-eye view of a plane defining a space (S21 ~ S35), assigning means (S37 to S55) for assigning a plurality of second areas respectively corresponding to the plurality of first areas designated by the assigning means to the object scene image output from the camera capturing the plane, and assigning by the assigning means Classifying means (S57 to S69, S75, S79) for classifying a plurality of selected second areas into suitability areas that match the size condition and improper areas that do not meet the size condition, and parts that belong to the appropriate area classified by the sort means First detection means (S83 to S93, S99 to S113) and classification means for detecting a motion coefficient (degree of congestion) of the scene (partial three-dimensional space) based on the scene image output from the camera Classified by Parts belonging to a proper area comprising a second detection means for detecting on the basis of motion coefficients of the object scene in motion coefficients detected by the first detecting means (S95 ~ S97, S115 ~ S131).

Preferably, the size condition corresponds to a condition that the size of the partial bird's-eye image representing the state of bird's-eye view of the partial scene belonging to the second area of interest exceeds the standard.

Preferably, a notification means (S71) for outputting a notification corresponding to the extraction of the inappropriate area by the classification means is further provided.

Preferably, the second detecting means designates each of one or more improper areas classified by the classifying means (S117, S131), and an appropriate area in contact with the improper area designated by the designation means. Search means (S121) for searching from one or more appropriate areas classified by the classification means, and the designation means based on the motion coefficient detected by the first detection means for the appropriate area found by the search means Calculation means (S125) for calculating the motion coefficient of the designated inappropriate area is included.

Preferably, adjustment means (S133) for adjusting the operations of the plurality of devices (D1 to D6) that generate outputs toward the space based on the detection results of the first detection means and / or the second detection means is further provided. .

More preferably, the designating means accepts a designating operation (S25 to S31) for designating each position of the plurality of devices, and defines a plurality of first areas based on the position designated by the designating operation. Means (S33).

The plurality of first areas are designated on a reference image representing a state in which the plane is viewed from the bird's eye, while the plurality of second areas are assigned to the object scene image output from the camera capturing the plane. For this reason, depending on the position and / or orientation of the camera, part or all of the second area may protrude from the object scene image, and the size of the second area allocated to the object scene image may be less than expected.

In the present invention, the plurality of second areas assigned to the object scene image are classified into an appropriate area that satisfies the size condition and an inappropriate area that deviates from the size condition, and the motion coefficient of the partial object scene belonging to the appropriate area is the camera. On the other hand, the motion coefficient of the partial scene belonging to the inappropriate area is detected based on the motion coefficient detected for the appropriate area. As a result, a decrease in motion detection accuracy due to a decrease in the size of the second area allocated to the object scene image is suppressed.

The above object, other objects, features, and advantages of the present invention will become more apparent from the following detailed description of embodiments with reference to the drawings.

It is a block diagram which shows the basic composition of one Example of this invention. It is a block diagram which shows the structure of one Example of this invention. It is an illustration figure which shows an example of the installation state of the camera applied to the FIG. 2 Example. It is an illustration figure which shows an example of the camera image displayed on the monitor of FIG. 2 Example. It is an illustration figure which shows an example of the map image displayed on the monitor of FIG. 2 Example. It is an illustration figure which shows an example of a structure of the area register applied to the FIG. 2 Example. FIG. 5 is an illustrative view showing one example of a configuration of a save register applied to the embodiment in FIG. 2; (A) is an illustration figure which shows an example of the allocation state of the divided area on a map image, (B) is an illustration figure which shows an example of the allocation state of the measurement area on a camera image. It is an illustration figure which shows a part of congestion degree measurement operation | movement. It is a flowchart which shows a part of operation | movement of CPU applied to the FIG. 2 Example. It is a flowchart which shows a part of other operation | movement of CPU applied to the FIG. 2 Example. FIG. 11 is a flowchart showing still another portion of behavior of the CPU applied to the embodiment in FIG. 2; FIG. 10 is a flowchart showing yet another portion of behavior of the CPU applied to the embodiment in FIG. 2; It is a flowchart which shows a part of other operation | movement of CPU applied to the FIG. 2 Example. FIG. 11 is a flowchart showing still another portion of behavior of the CPU applied to the embodiment in FIG. 2; FIG. 10 is a flowchart showing yet another portion of behavior of the CPU applied to the embodiment in FIG. 2; It is a flowchart which shows a part of other operation | movement of CPU applied to the FIG. 2 Example.

Embodiments of the present invention will be described below with reference to the drawings.
[Basic configuration]
Referring to FIG. 1, the motion detection apparatus of this embodiment is basically configured as follows. The designation unit 1 designates a plurality of first areas on a reference image that represents a bird's eye view of a plane that defines a space. The assigning unit 2 assigns a plurality of second areas respectively corresponding to the plurality of first areas specified by the specifying unit 1 to the object scene image output from the camera 6 capturing a plane. The classifying unit 3 classifies the plurality of second areas assigned by the assigning unit 2 into suitability areas that match the size condition and improper areas that fall outside the size condition. The first detection means 4 detects the motion coefficient of the partial scene belonging to the appropriate area classified by the classification means 3 based on the scene image output from the camera 6. The second detection means 5 detects the motion coefficient of the partial scene belonging to the inappropriate area classified by the classification means 3 based on the motion coefficient detected by the first detection means 4.

The plurality of first areas are designated on a reference image representing a state in which the plane is viewed from the bird's eye, while the plurality of second areas are assigned to the object scene image output from the camera 6 capturing the plane. For this reason, depending on the position and / or orientation of the camera 6, part or all of the second area may protrude from the object scene image, and the size of the second area allocated to the object scene image may be less than expected. .

In this embodiment, the plurality of second areas allocated to the object scene image are classified into appropriate areas that satisfy the size condition and inappropriate areas that deviate from the size condition, and the motion coefficient of the partial object scene belonging to the appropriate area is While detected based on the object scene image output from the camera 6, the motion coefficient of the partial object scene belonging to the inappropriate area is detected based on the motion coefficient detected for the appropriate area. As a result, a decrease in motion detection accuracy due to a decrease in the size of the second area allocated to the object scene image is suppressed.
[Example]
Referring to FIG. 2, the congestion degree measuring apparatus 10 of this embodiment includes a camera 12 that repeatedly outputs image data representing an object scene (three-dimensional space) captured on an imaging surface. The image data output from the camera 12 is captured by the image processing circuit 14 and subjected to camera image display processing by the CPU 14p. As a result, an image representing the object scene, that is, a camera image is displayed on the monitor 16.

Referring to FIG. 3, the camera 12 is installed on the upper surface of the wall of the room having the plane FS1, and captures the plane FS1 obliquely from above. Therefore, the camera image is displayed on the monitor screen as shown in FIG. As shown in FIGS. 3 and 4, the plane FS1 is defined by the X axis and the Y axis orthogonal to each other, and the camera image is reproduced along the U axis and the V axis orthogonal to each other.

Air conditioners D_1 to D_6 are installed at a predetermined distance on the ceiling of the room. Each of the air conditioners D_1 to D_6 has four outlets respectively corresponding to a total of four directions including two directions along the X axis and two directions along the Y axis, and outputs air having a specified temperature with a specified air volume. .
The room temperature is adjusted by the air thus output.

When the measurement area setting mode is selected by operating the input device 18, the following processing is executed by the CPU 14p.

First, the map image shown in FIG. A map image is corresponded to the image which represents the state which looked at plane FS1 bird's-eye view. In the map image, marks M_1 to M_6 respectively representing air conditioners D_1 to D_6 are displayed corresponding to the positions of the air conditioners D_1 to D_6.

When the display of the map image is completed, the variable K is set to “1”. When the mark M_K is clicked by the mouse pointer provided in the input device 18, coordinates indicating the clicked position, that is, click coordinates are calculated. The calculated click coordinates are described in the area register 14r1 corresponding to the variable K, and the variable K is incremented thereafter. The click operation is accepted a total of six times corresponding to the marks M_1 to M_6, and thereby six click coordinates respectively corresponding to the marks M_1 to M_6 are set in the area register 14r1.

The map image is divided in the manner shown in FIG. 8A with the six click coordinates thus set as a reference. A thick line and a thin line that define a division position are drawn on the map image. The thick lines are drawn so as to surround the marks M_1 to M_6, and the thin lines are drawn so as to extend radially from each of the marks M_1 to M_6.

As a result, the divided areas MP_1_1 to MP_1_4 are allocated around the mark M_1, the divided areas MP_2_1 to MP_2_4 are allocated around the mark M_2, and the divided areas MP_3_1 to MP_3_4 are allocated around the mark M_3. Similarly, the divided areas MP_4_1 to MP_4_4 are allocated around the mark M_4, the divided areas MP_5_1 to MP_5_4 are allocated around the mark M_5, and the divided areas MP_6_1 to MP_6_4 are allocated around the mark M_6.

In the area register 14r1, a plurality of XY coordinates defining the divided area MP_K_L (K: 1 to 6, L: 1 to 4) allocated in this way, and a normalization coefficient α_K_L indicating a numerical value corresponding to the area of the divided area MP_K_L Is described. The normalization coefficient α_K_L is obtained by dividing the area of the divided area MP_K_L by the unit area.

Subsequently, each of a plurality of XY coordinates that define the divided area MP_K_L is converted into UV coordinates according to Equation 1.

The calibration parameters P11 to P33 shown in Equation 1 correspond to a matrix for performing planar projective transformation between the XY coordinate system that defines the plane FS1 and the UV coordinate system that defines the camera image. Therefore, by applying the desired XY coordinates to Equation 1, the corresponding UV coordinates on the camera image are calculated. The plurality of UV coordinates thus converted are described in the area register 14r1 corresponding to the plurality of XY coordinates of the conversion source, and the measurement area DT_K_L corresponding to the divided area MP_K_L is displayed on the camera image as shown in FIG. Defined.

When the definition of the measurement area DT_K_L is completed, the camera image belonging to the defined measurement area DT_K_L is converted into a bird's-eye image, and the size of the converted bird's-eye image is detected as “SZ_K_L”. The bird's-eye view image corresponds to an image representing a state where the plane FS1 is captured from the virtual camera arranged at a predetermined height, and the size SZ_K_L indicates a numerical value normalized by the height of the virtual camera. The size SZ_K_L detected in this way is described in the area register 14r1 corresponding to the variables K and L.

When a part of the measurement area DT_K_L protrudes from the camera image, the size SZ_K_L has a bird's-eye view generated based on a part of the camera images belonging to another part of the measurement area DT_K_L (measurement area within the camera image). Indicates the size of the image. In FIG. 8B, for example, a part of the measurement area DT_1_1 protrudes from the camera image. In this case, the size SZ_1_1 indicates the size of the bird's-eye view image based on the camera image in the area indicated by hatching in FIG.

The measurement area DT_K_L defined in this way is classified into an appropriate area that matches the size condition and an inappropriate area that deviates from the size condition. Here, the size condition corresponds to a condition that the size SZ_K_L exceeds the reference value REF. Therefore, the measurement area DT_K_L in which the size SZ_K_L exceeds the reference value REF is classified as an appropriate area, and the measurement area DT_K_L in which the size SZ_K_L is equal to or smaller than the reference value REF is classified as an inappropriate area.

When an inappropriate area is detected, an error is notified. The error is notified in a state where the inappropriate area is specified, such as adding another color to the inappropriate area. The operator recognizes the presence of the inappropriate area by such notification.

When the operator performs a cancel operation on the input device 18 in response to the error notification, the area register 14r1 is cleared. The operator changes the arrangement of the camera 12 and / or the air conditioners D1 to D6, and then performs the operation under the measurement area setting mode again. On the other hand, when the operator performs a registration operation on the input device 18 in response to the error notification, an extension indicating that the measurement area DT_K_L is an inappropriate area is described in the area register 14r1 corresponding to the variables K and L. The The extension set in the measurement area DT_K_L is referred to in the congestion degree measurement mode.

When the congestion degree measurement mode is selected by operating the input device 18, the following processing is executed by the CPU 14p every time the measurement period arrives.

First, an image showing motion, that is, a motion image is detected on the camera image. Subsequently, the flag FLG is set to “0”, and it is determined whether or not an extension is set in the measurement area DT_K_L (that is, whether or not the measurement area DT_K_L is an appropriate area).

If no extension is set, the measurement area DT_K_L is regarded as an appropriate area, and the number of pixels of the camera image belonging to the measurement area DT_K_L is detected. The number of detected pixels is set in a variable P1_K_L. In addition, a motion image belonging to the measurement area DT_K_L is segmented from the motion image detected on the camera image, and the number of pixels of the segmented motion image is set to the variable P2_K_L.

The congestion degree CR_K_L indicating the degree of congestion of the measurement area DT_K_L is calculated based on the variables P1_K_L and P2_K_L set in this way and the normalization coefficient α_K_L described in the register 14r. Specifically, the degree of congestion CR_K_L is obtained by dividing the variable P2_K_L by the variable P1_K_L and multiplying the division value obtained thereby by the normalization coefficient α_K.

On the other hand, if an extension is set in the measurement area DT_K_L, the measurement area DT_K_L is regarded as an inappropriate area, and the values of the variables K and L are described in the common column of the save register 14r2 shown in FIG. Further, the flag FLG is updated to “1” in order to assert that the values of the variables K and L have been saved.

不 For inappropriate areas, the congestion level is calculated as follows. First, the variable N is set to each of “1” to “Nmax” (Nmax: total number of columns in which numerical values are described in the save register 14r2), and the two numerical values described in the Nth column of the save register 14r2 are variables. Set to K and L, respectively. As a result, one or more improper areas are designated in order.

Subsequently, one or more appropriate areas adjacent to the designated inappropriate area are searched from the periphery of the designated inappropriate area. When one or more appropriate areas adjacent to the specified inappropriate area are found, the average value of the congestion levels of the detected one or more appropriate areas is set as the congestion level of the specified inappropriate area. Is done. On the other hand, if a proper area adjacent to the designated inappropriate area is not found, a default value is set as the congestion degree of the designated inappropriate area.

The outputs of the air conditioners D1 to D6 are controlled based on the congestion levels CR_1_1 to CR_6_4 thus obtained. Specifically, the output of the air conditioner corresponding to the measurement area with a high degree of congestion is strengthened, and the output of the air conditioner corresponding to the measurement area with a low degree of congestion is weakened.

Referring to FIG. 9, when measurement areas DT_1_2 and DT_1_3 are appropriate areas, but measurement areas DT_1_1 and DT_1_4 are inappropriate areas, persons H1 and H2 move measurement area DT_1_2, and person H3 moves to measurement area. Assume that DT_1_3 is moved.

Then, the degree of congestion of the measurement area DT_1_2 is calculated based on the number of pixels of the image belonging to the measurement area_1_2, the number of pixels of the motion image representing the persons H1 and H2, and the normalization coefficient α_1_2, and the degree of congestion of the measurement area DT_1_3 is measured. It is calculated based on the number of pixels of the image belonging to the area DT_1_3, the number of pixels of the motion image representing the person H3, and the normalization coefficient α_1_3. On the other hand, the congestion degree of the measurement area DT_1_1 is calculated based on the congestion degree of the measurement area DT_1_2, and the congestion degree of the measurement area DT_1_4 is calculated based on the congestion degree of the measurement area DT_1_3.

The CPU 14p executes a plurality of tasks including an imaging task shown in FIG. 10, a measurement area setting task shown in FIGS. 11 to 14, and a congestion degree measuring task shown in FIGS. Note that control programs corresponding to these tasks are stored in the recording medium 20.

Referring to FIG. 10, camera image display processing is executed in step S1. As a result, a camera image is displayed on the monitor 16. In step S3, it is determined whether or not the current operation mode is the measurement area setting mode, and in step S7, it is determined whether or not the current operation mode is the congestion degree measurement mode.

If “YES” in the step S3, a measurement area setting task is started in a step S5, and thereafter, the process proceeds to a step S15. If “YES” in the step S7, it is determined whether or not the measurement area has been set in a step S9. If the determination result is YES, the congestion degree measurement task is started in step S11 and then the process proceeds to step S15. If the determination result is NO, the process directly proceeds to step S15. If both step S3 and S7 are NO, another process is executed in step S13, and then the process proceeds to step S15.

In step S15, it is repeatedly determined whether or not a mode change operation has been performed. When the determination result is updated from NO to YES, the activated task is terminated in step S17, and thereafter, the process returns to step S3.

Referring to FIG. 11, the map image is displayed on monitor 16 in step S21, and variable K is set to “1” in step S23. In step S25, it is determined whether or not a click operation for designating an area has been performed. If the determination result is updated from NO to YES, click coordinates are calculated in step S27. The calculated coordinates are described in the area register 14r1 corresponding to the variable K. In step S29, it is determined whether or not the variable K has reached “6”. If the determination result is NO, the variable K is incremented in step S31 and then the process returns to step S25. Proceed to S33.

In step S33, the map image is divided with reference to the click coordinates. As a result, a total of 24 divided areas MP_1_1 to MP_6_4 are allocated on the map image. In step S35, bold and thin lines indicating the positions of the divided images MP_1_1 to MP_6_4 are drawn on the map image.

In step S37, the variable K is set to “1”, in step S39, the variable L is set to “1”, and in step S41, a plurality of XY coordinates defining the divided area MP_K_L are calculated. The calculated XY coordinates are described in the area register 14r1 corresponding to the variables K and L. In step S43, the area of the divided area MP_K_L is normalized, and a normalization coefficient α_K_L is calculated. The calculated normalization coefficient α_K_L is also described in the area register 14r1 corresponding to the variables K and L.

In step S45, each of the plurality of XY coordinates defining the divided area MP_K_L is converted into UV coordinates according to the equation (1). The converted UV coordinates are described in the area register 14r1 corresponding to the variables K and L, whereby the measurement area DT_K_L corresponding to the divided area MP_K_L is assigned to the camera image. Depending on the position and / or orientation of the camera 12, at least part of the measurement area DT_K_L may protrude from the camera image.

In step S47, a part of the camera images belonging to the measurement area DT_K_L allocated in this way is converted into a bird's-eye view image, and the size of the converted bird's-eye view image is detected as “SZ_K_L”. The detected size SZ_K_L is described in the area register 14r1 corresponding to the variables K and L.

In step S49, it is determined whether or not the variable L has reached “4”, and in step S53, it is determined whether or not the variable K has reached “6”. If the decision result in the step S49 is NO, the variable L is incremented in a step S51, and then the process returns to the step S41. If the determination result in step S49 is YES and the determination result in step S53 is NO, the variable K is incremented in step S55, and the process returns to step S39. If both the determination result in step S49 and the determination result in step S53 are YES, the process proceeds to step S57.

In step S57, the variable K is set to “1”, in step S59, the variable L is set to “1”, and in step S61, it is determined whether or not the size SZ_K_L exceeds the reference value REF. If the determination result is NO, the process proceeds to step S71, and if the determination result is YES, the process proceeds to step S63. As a result, the measurement area DT_K_L whose size SZ_K_L exceeds the reference value REF is classified as an appropriate area, and the measurement area DT_K_L whose size SZ_K_L is equal to or smaller than the reference value REF is classified as an inappropriate area.

In step S63, it is determined whether or not the variable L has reached “4”. In step S67, it is determined whether or not the variable K has reached “6”. If the decision result in the step S63 is NO, the variable L is incremented in a step S65, and then the process returns to the step S61. If the determination result in step S63 is YES and the determination result in step S67 is NO, the variable K is incremented in step S69, and the process returns to step S59. If both the determination result in step S63 and the determination result in step S67 are YES, the process ends.

In step S71, an error is notified in a state where the measurement area DT_K_L is specified. In step S73, it is determined whether a cancel operation has been performed. In step S75, it is determined whether a registration operation has been performed. If the decision result in the step S73 is YES, the area register 14r1 is cleared in a step S77, and then the process is ended. If the decision result in the step S75 is YES, the process advances to a step S79 to describe the extension in the area register 14r1 corresponding to the variables K and L. When the description of the extension is completed, the process returns to step S63.

Referring to FIG. 15, it is determined in step S81 whether or not a measurement cycle has arrived. When the determination result is updated from NO to YES, the process proceeds to step S83, and an image showing motion, that is, a motion image is detected on the camera image. In step S85, the flag FLG is set to “0”, in step S87, the variable K is set to “1”, and in step S89, the variable L is set to “1”. In step S91, it is determined whether or not an extension is assigned to the measurement area DT_K_L with reference to the area register 14r1. If the determination result is NO, the process proceeds to step S93. If the determination result is YES, the process proceeds to step S95. move on.

In step S95, the current values of the variables K and L are set in the save register 14r2, and in step S97, the flag FLG is set to “1”. When the process of step S97 is completed, the process proceeds to step S107.

In step S93, the number of pixels of the camera image belonging to the measurement area DT_K_L is detected, and the detected number of pixels is set in the variable P1_K_L. In step S99, a motion image belonging to the measurement area DT_K_L is segmented from the motion image detected in step S83, and in step S101, the number of pixels of the segmented motion image is set to a variable P2_K_L.

In step S103, the ratio of the motion image belonging to the measurement area DT_K_L to the measurement area DT_K_L is calculated based on the variables P1_K_L and P2_K_L set in steps S93 and S101, respectively. The ratio (= RT_K_L) corresponds to a value obtained by dividing the variable P2_K_L by the variable P1_K_L. In step S105, the normalization coefficient α_K_L described in the area register 14r1 is multiplied by the ratio RT_K_L calculated in step S103 to calculate the congestion degree of the measurement area DT_K_L as “CR_K_L”.

In step S107, it is determined whether or not the variable L has reached “4”, and in step S111, it is determined whether or not the variable K has reached “6”. If the determination result in step S107 is NO, the variable L is incremented in step S109, and the process returns to step S91. If the determination result in step S107 is YES and the determination result in step S111 is NO, the variable K is incremented in step S113, and the process returns to step S89. If both the determination result in step S107 and the determination result in step S111 are YES, the process proceeds to step S115.

In step S115, it is determined whether or not the flag FLG is “1”. If the determination result is NO, the process proceeds to step S133, and if the determination result is YES, the process proceeds to step S117. In step S117, the variable N is set to “1”, and in step S119, the two numerical values set in the Nth column of the save register 14r2 are set in the variables K and L, respectively.

In step S121, one or two or more appropriate areas are searched from the periphery of the measurement area DT_K_L. In step S123, it is determined whether or not the search is successful, that is, whether one or more adjacent appropriate areas are found.

If the determination result is YES, the process proceeds to step S129 through the process of step S125, and if the determination result is NO, the process proceeds to step S129 through the process of step S127. In step S125, an average value of one or more congestion levels corresponding to one or more found appropriate areas is calculated as the congestion level CR_K_L of the measurement area DT_K_L. In step S127, a default value is set as the congestion degree CR_K_L.

In step S129, it is determined whether or not the variable N has reached the maximum value Nmax. If the determination result is NO, the variable N is incremented in step S131, and the process returns to step S119. If the determination result is YES, step S129 is performed. The process proceeds to S133. In step S133, the outputs of the air conditioners D1 to D6 are controlled based on the degree of congestion obtained in steps S105, S125, and S127. When such air conditioning control is completed, the process returns to step S81.

As can be seen from the above description, the CPU 14p designates a plurality of divided areas MP_1_1 to MP_6_4 on the map image representing the bird's-eye view of the plane FS1 defining the space (S21 to S35), and designates the plurality of designated divided areas. A plurality of measurement areas DT_1_1 to DT_6_4 respectively corresponding to MP_1_1 to MP_6_4 are allocated to the object scene image output from the camera 12 capturing the plane FS1 (S37 to S55). The CPU 14p also classifies the measurement areas DT_1_1 to DT_6_4 assigned to the camera image into appropriate areas that match the size conditions and inappropriate areas that are out of the size conditions (S57 to S69, S75, S79), and parts that belong to the appropriate areas The degree of congestion (= motion coefficient) of the object scene is detected based on the object scene image output from the camera 12 (S83 to S93, S99 to S113), while the degree of congestion of the partial object scene belonging to the inappropriate area Is detected based on the degree of congestion detected for the appropriate area (S95 to S97, S115 to S131).

The divided areas MP_1_1 to MP_6_4 are specified on a map image representing a state in which the plane FS1 is viewed, while the measurement areas DT_1_1 to DT_6_4 are assigned to the object scene image output from the camera 12 capturing the plane FS1. Therefore, depending on the position and / or orientation of the camera 12, some or all of the measurement area may protrude from the object scene image, and the size of the measurement area remaining on the object scene image may be less than expected. There is.

In this embodiment, the measurement areas DT_1_1 to DT_6_4 are classified into an appropriate area that satisfies the size condition and an inappropriate area that deviates from the size condition, and the degree of congestion of the partial field belonging to the appropriate area is determined by the object output from the camera 12. While detected based on the field image, the degree of congestion of the partial object scene belonging to the inappropriate area is detected based on the degree of congestion detected for the appropriate area. As a result, a decrease in motion detection accuracy due to a decrease in the size of the measurement area assigned to the object scene image is suppressed.

In this embodiment, the coordinates of the marks M_1 to M_6 are designated by clicking the mouse pointer. However, instead of this, the coordinate values of the marks M_1 to M_6 may be directly specified.

In this embodiment, the plane FS1 is captured obliquely from above by the camera 12 installed on the upper surface of the wall. Instead, the plane FS1 is captured from directly above by the omnidirectional camera set on the ceiling. It may be.

Further, in this embodiment, it is assumed that the output of the air conditioner is adaptively controlled. However, instead of the output of the air conditioner or together with the output of the air conditioner, the output (that is, brightness) of the lighting device is adaptive. You may make it control to.

In this embodiment, the planar projective transformation referring to Equation 1 is assumed, but a perspective projective transformation may be performed instead.

In this embodiment, an image schematically representing a state in which the plane FS1 is viewed from the bird's-eye view is employed as the map image. However, the map image may be generated by performing bird's-eye conversion with reference to Equation 1 described above on the camera image. However, in this case, a part of the plane FS1 may not be reproduced depending on the position and / or orientation of the camera 12.

DESCRIPTION OF SYMBOLS 10 ... Congestion degree measuring device 12 ... Camera 14 ... Image processing circuit 14p ... CPU
16 ... monitor 18 ... input device

Claims

A designation means for designating a plurality of first areas on a reference image representing a state in which a plane defining the space is bird's-eye view;
Assigning means for assigning a plurality of second areas respectively corresponding to the plurality of first areas designated by the designation means to a scene image output from a camera capturing the plane;
Classifying means for classifying the plurality of second areas assigned by the assigning means into suitability areas that match a size condition and improper areas that fall outside the size condition;
First detection means for detecting a motion coefficient of a partial scene belonging to an appropriate area classified by the classification means based on a scene image output from the camera; and an inappropriate area classified by the classification means A motion detection apparatus comprising: second detection means for detecting a motion coefficient of a partial object scene belonging to the first scene based on the motion coefficient detected by the first detection means.
The motion detection apparatus according to claim 1, wherein the size condition corresponds to a condition that a size of a partial bird's-eye view image representing a state in which the partial scene belonging to the second area of interest exceeds a reference.
The motion detection apparatus according to claim 1 or 2, further comprising notification means for outputting a notification corresponding to the extraction of the inappropriate area by the classification means.
The second detecting means designates each of one or more improper areas classified by the classifying means, and classifies an appropriate area in contact with the improper area designated by the designation means by the classifying means. Search means for searching from among the one or more appropriate areas that have been specified, and the error specified by the specifying means based on the motion coefficient detected by the first detection means for the appropriate area found by the search means. The motion detection apparatus according to claim 1, further comprising calculation means for calculating a motion coefficient of an appropriate area.
5. The apparatus according to claim 1, further comprising an adjustment unit that adjusts operations of a plurality of devices that generate an output toward the space based on a detection result of the first detection unit and / or the second detection unit. The motion detection device described.
The designation unit includes a reception unit that receives a designation operation that designates the position of each of the plurality of devices, and a definition unit that defines the plurality of first areas on the basis of the position designated by the designation operation. Item 6. The motion detection device according to Item 5.