JP2004150814A - Method and apparatus for adjusting imaging position - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a method for adjusting an imaging position capable of stably determining the imaging position even in the case of the presence of variations in a region in an image and in the degree of detection. <P>SOLUTION: A digital image is inputted, an object to be recognized is extracted, and flags are placed on pixels at which the object to be recognized exists in the image. Coordinates values of the pixels at which the flags are placed are added. The added values are divided by the pixel number of the pixels at which the flags are placed. The position of the center of gravity of the object to be recognized is determined according to this result. The position of moment of a camera is computed on the basis of the center of gravity. <P>COPYRIGHT: (C)2004,JPO

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to an imaging method and an imaging device.
[0002]
[Prior art]
Japanese Patent Application Laid-Open No. 2000-147792 discloses a technique for appropriately paying attention to a face or the like from an image captured by an imaging unit. In this technique, an edge having the characteristics of the upper eyelid of a human eye (eyelid) is extracted from a face image, and eyes are detected from the face image based on the edge.
[0003]
[Patent Document 1]
JP-A-2000-137792 (pages 3 to 4, FIG. 4)
[0004]
[Problems to be solved by the invention]
By the way, when an object is extracted from an image based on a predetermined feature as in the above-described related art, when there are a plurality of objects in the image, there is a problem as to where to place the point of interest. When there are a plurality of objects, the camera is directed to an object having a large area or an object having a high degree of detection. For this purpose, a point of interest in a plurality of objects is calculated by calculating the area occupied by the object in the image or the degree of detection of the object as a weighting factor and multiplying the weight on the detected position on the image. I was Alternatively, attention is paid to the center of gravity of each of a plurality of objects in an image, and the point of interest is determined by integrating these centers of gravity.
[0005]
However, in the case of a method of performing weighting in accordance with the size and detection degree of the area occupied by each of a plurality of objects, due to area fluctuation or noise when the objects move and the objects are merged in an image. When the detection degree fluctuates, there is a problem that the attention point cannot be accurately calculated in some cases. In addition, the method that focuses on the center of gravity of a plurality of objects is easy to stabilize with respect to the variation in the area and the degree of detection, but requires a complicated calculation when calculating the center of gravity of each object and integrating them. There was a problem that it would take.
[0006]
Therefore, an object of the present invention is to provide an imaging position adjustment method and an imaging position adjustment method that can stably determine an imaging position even when there is an area variation or detection degree variation in an image, and does not require complicated arithmetic processing. It is to provide a position adjusting device.
[0007]
[Means for Solving the Problems]
An imaging position adjusting method according to the present invention that solves the above problem extracts a recognition target in an image captured by an imaging unit, and adjusts an imaging position of the imaging unit so that the recognition target exists in the image. This is a method of adjusting an imaging position, in which a specific position in an image is obtained by performing weighting according to the area of a recognition target in an image, and an imaging position of an imaging unit is determined based on the specific position.
[0008]
In the imaging position adjustment method according to the present invention, when there are a plurality of recognition targets, instead of treating those targets individually, weighting is performed according to the area of the recognition target in the image, and Seeking a specific location. For this reason, even if there is a region variation or a detection degree variation in the image, the influence can be reduced, and the imaging position can be stabilized.
[0009]
Here, when performing the weighting, it is preferable that the specific position is obtained by performing the weighting in consideration of the existence time of the recognition target in the image.
[0010]
In the imaging position adjustment method according to the present invention, when performing weighting for obtaining a specific position, weighting is performed in consideration of the existence time of the recognition target object. For this reason, even if the recognition target moves due to a change over time, the influence can be reduced, and the imaging position can be further stabilized.
[0011]
In addition, the imaging position adjusting method according to the present invention that has solved the above-described problem extracts a recognition target in an image captured by the imaging unit, and adjusts an imaging position of the imaging unit so that the recognition target exists in the image. An imaging position adjustment method for adjusting, wherein a coordinate value of an area where an object to be recognized is present in an image is added, and a value obtained by adding the coordinate values is divided by the number of areas where the object to be recognized is present. The position corresponding to the value is determined as a specific position in the image, and the imaging position of the imaging unit is determined based on the specific position.
[0012]
In the imaging position adjustment method according to the present invention, the area where the recognition target exists in the image, for example, the coordinate values of the pixels are added, and the added value is divided by the number of pixels where the recognition target exists, and the specific position is determined. Seeking. As described above, since the center of gravity of the position of the recognition target on the entire surface of the image is determined as the specific position, it is possible to stably determine the imaging position even when there is a region variation or a detection degree variation in the image. Can be. In addition, in the calculation for obtaining the center of gravity, the division is performed after adding the coordinate values of all the pixels. As described above, since the division requiring a long time for the processing can be performed only once, complicated arithmetic processing is unnecessary, and as a result, the arithmetic time can be shortened.
[0013]
At this time, the addition of the coordinate values is repeatedly performed a plurality of times during a predetermined time, and when a specific position in the image is obtained, the value obtained by adding the coordinate values during the predetermined time is determined during the predetermined time. It is preferable to obtain a position corresponding to a value obtained by dividing the area where the recognition target object exists by the number of times of addition as the specific position in the image.
[0014]
As described above, by adding the coordinate values a plurality of times repeatedly for a predetermined time in consideration of the time change in the addition of the coordinate values, even if the recognition target moves due to the time change, etc. The influence can be reduced, and the imaging position can be further stabilized.
[0015]
Further, the imaging position adjusting method according to the present invention that has solved the above-described problem extracts a recognition target in an image captured by the imaging unit, and adjusts an imaging position of the imaging unit so that the recognition target exists in the image. An imaging position adjustment method for adjusting, wherein weighting is performed on a coordinate value of an area that satisfies a predetermined condition, a coordinate value of an area that satisfies the weighted predetermined condition is added, and the coordinate value is added. A position corresponding to a value obtained by dividing the obtained value by the number of regions satisfying a predetermined condition is determined as a specific position in the image, and the imaging position of the imaging unit is determined based on the specific position.
[0016]
In the imaging position adjustment method according to the present invention, weighting is performed on coordinate values of an area that satisfies a predetermined condition, such as a skin color when a face is detected, and the weighted predetermined condition is determined. The coordinate values of the area to be filled are added. A position that satisfies the predetermined condition can be considered as a position where the recognition target object exists. The center of gravity is calculated by dividing the obtained sum of the coordinates of the region satisfying the predetermined condition by the number of the regions, and the center of gravity is determined as the specific position. As described above, since the center of gravity of the position of the recognition target on the entire surface of the image is set as the specific position, the imaging position can be determined stably even when there is a region variation or a detection degree variation in the image. . In addition, in the calculation for obtaining the center of gravity, the division is performed after adding the coordinate values of all the pixels. As described above, since a division requiring a long processing time can be performed only once, complicated calculation processing is not required, and as a result, the calculation time can be shortened.
[0017]
Here, the weighting of the coordinate values is repeatedly performed a plurality of times during a predetermined time, and when a specific position in an image is obtained, a value obtained by adding the coordinate values satisfies a predetermined condition during a predetermined time. It is preferable to obtain a position corresponding to a value obtained by dividing the number of times of adding the area as a specific position in the image.
[0018]
In this way, when adding the coordinate values, by adding the coordinate values for a predetermined time in consideration of the time change, even if the recognition target moves due to the time change, the influence thereof is reduced. Therefore, the imaging position can be further stabilized.
[0019]
In addition, an imaging position adjusting apparatus according to the present invention that solves the above-described problems includes an image processing unit that extracts a recognition target in an image captured by the imaging unit, and an imaging unit that detects the recognition target in the image. Imaging position control means for performing control for adjusting the imaging position of the image, wherein the image processing means includes a coordinate value adding means for adding a coordinate value of a region where the recognition target exists in the image, and a coordinate value means. Specific position calculating means for obtaining a position corresponding to a value obtained by dividing the obtained value by the number of regions where the recognition target object is present as a specific position in the image; and an imaging position for determining an imaging position of the imaging means based on the specific position. Determination means.
[0020]
Here, the coordinate value adding means calculates the coordinate value of the area where the recognition target object exists for a predetermined time when adding the coordinate values, and the specific position calculating means calculates the specific position in the image. A position corresponding to a value obtained by dividing the value obtained by the coordinate value adding means during the predetermined time by the number of times of adding the area where the recognition target object exists during the predetermined time is obtained as a specific position in the image. Is preferred.
[0021]
Further, the imaging position adjusting apparatus according to the present invention, which has solved the above-mentioned problems, comprises: an image processing means for extracting a recognition target in an image picked up by the imaging means; and an imaging means so that the recognition target exists in the image. Imaging position control means for performing control to adjust the imaging position of the image processing apparatus, wherein the image processing means weights coordinate values of an area satisfying a predetermined condition, and a weighted predetermined value. A coordinate value adding means for adding the coordinate values of the area satisfying the condition, and a position in the image in which the value obtained by dividing the value obtained by the coordinate value adding means by the number of areas satisfying the predetermined condition is specified. It has a specific position calculating means for obtaining a position, and an imaging position determining means for determining an imaging position of the imaging means based on the specific position.
[0022]
At this time, the weighting means weights the coordinate values of an area satisfying a predetermined condition for a predetermined time when performing the weighting, and the specific position calculating means determines when the specific position in the image is obtained. Preferably, a position corresponding to a value obtained by dividing the value obtained by adding the coordinate values by the number of times of adding the region satisfying the predetermined condition during the predetermined time is obtained as the specific position in the image.
[0023]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the drawings.
[0024]
FIG. 1 is a block diagram of an imaging device including an imaging position adjustment device according to an embodiment of the present invention.
[0025]
As shown in FIG. 1, an imaging apparatus M according to the present embodiment includes a camera 1 as an imaging unit and a driving device 2 as an imaging unit driving unit that moves the camera 1 to adjust an imaging position of the camera 1. Have. The driving device 2 includes a base 11 and a stage 12. The base 11 is fixed on the ground, for example, and the stage 12 is attached to the base 11 so as to be rotatable around a vertical axis and a horizontal axis. The camera 1 is mounted on the stage 12 of the driving device 2, and the imaging position of the camera 1 can be freely adjusted by driving the stage 12.
[0026]
Further, the imaging device M includes a recognition control device 3 which is an imaging position adjustment device according to the present invention, and both the camera 1 and the driving device 2 are connected to the recognition control device 3. The recognition control device 3 includes an A / D conversion unit 21, an image processing unit 22 as an image processing unit, and a drive control unit 23 as an imaging position control unit. The camera 1 provided outside the recognition control device 3 is connected to the A / D conversion unit 21, and the drive device 2 is connected to the drive control unit 23.
[0027]
From the camera 1, an image signal of an analog image captured by the camera 1 is output to the A / D converter 21. The A / D converter 21 converts the analog image signal output from the camera 1 into a digital image by A / D conversion. The A / D converter 21 outputs the image signal of the digital image to the image processor 22.
[0028]
The image processing unit 22 calculates an image pickup position of the camera 1 by extracting a recognition target in the image, for example, a human face, from the digital signal output from the A / D conversion unit 21 and performing predetermined image processing on the object. I do. Based on the imaging position, the movement amount of the camera 1, that is, the movement amount of the stage 12 in the driving device 2 is calculated.
[0029]
The image processing unit 22 includes a memory (not shown), and can store whether or not a flag is set for each pixel in the image. The image processing apparatus also includes a calculation unit (not shown), and performs an addition process of adding the coordinate values of each pixel in the image and a division process of dividing a value obtained by adding the coordinate values by the number of pixels. be able to. Then, it is possible to perform a process of determining the imaging position of the camera 1 based on the specific position obtained by the division process. This calculation unit constitutes a coordinate adding unit, a specific position calculating unit, and an imaging position determining unit of the present invention. The details of the image processing performed by the image processing unit 22 will be described later.
[0030]
The movement amount of the stage 12 calculated by the image processing unit 22 is output to the drive control unit 23. The drive control unit 23 generates a control signal for controlling the stage 12 of the driving device 2 based on the imaging position adjustment signal output from the image processing unit 22, and outputs the control signal to the driving device 2. The drive device 2 drives the stage 12 based on the control signal output from the drive control unit 23 of the recognition control device 3 to adjust the imaging position of the camera 1.
[0031]
Next, the procedure for adjusting the imaging position according to the present embodiment will be described.
[0032]
The adjustment of the imaging position is performed based on the image captured by the camera 1. An analog image captured by the camera 1 and output from the camera 1 to the A / D conversion unit 21 of the recognition control device 3 is converted into a digital image and output to the image processing unit 22. Then, the image processing unit 22 processes the digital image and adjusts the imaging position of the camera according to the flow shown in FIG. FIG. 2 is a flowchart showing the procedure of the imaging position adjustment method according to the present invention.
[0033]
First, a digital image output from the A / D converter 21 is input (S1). At this time, among the pixels, the voting memory remains at the pixel for which the flag has been set by the previous processing, so that the voting memory is cleared to 0 in order to perform the current processing (S2). After the voting memory is cleared to 0, the process of extracting the recognition target is performed (S3). When the recognition target is extracted, a feature of the recognition target that satisfies a predetermined condition is used. Features that can be used here include color, brightness, brightness, and the like. For example, when the object to be recognized is a human face, the feature is a skin color or a color close to the skin color, and a pixel having this feature can be a detection target.
[0034]
Specifically, it is assumed that human faces F1, F2, and F3 exist in the image 30, as shown in FIG. In this case, as shown in FIG. 3B, the human face in the image 30 is considered to be in the regions E1, E2, and E3, respectively. At this time, whether or not the face is a human face can be determined based on, for example, whether or not the color is a skin color or a color close to the skin color. Then, after defining the regions E1 to E3 where the human face is located, as shown in FIG. 3C, the image of the image 30 is captured in units of pixels P, P... It is determined whether or not it is included. The positions of these pixels are shown as points on the XY coordinates.
[0035]
In this way, as a process of extracting a target object, if it is determined whether or not the image is a human face, a flag is set for a pixel including an area determined as a human face (hereinafter, referred to as a “face area”) (S4). . Thus, by setting a flag on a pixel including a face area, it is possible to determine whether or not the pixel includes a face area. Then, voting is performed for the pixel on which the flag is set, and the coordinate value is added (S5). Then, the result of adding the coordinate values is divided by the total number of pixels for which voting has been performed (S6). By performing such processing, it is possible to obtain the center of gravity that is the specific position of the face area in the image (S7). By setting the specific position as the center of gravity, a result obtained by performing weighting according to the area of the recognition target object can be set as the specific position.
[0036]
In obtaining the center of gravity of the face region in the image in this manner, the processes of steps S4 to S6 are the same as the processes of calculating the first moment of the image. Therefore, the center of gravity of the face area can be obtained by the following equations (1) and (2).
[0037]
Mx = ∫x · f (x, y) dx (1)
My = ∫y · f (x, y) dy (2)
[0038]
After the center of gravity of the face area is obtained in this way, the movement position of the camera 1 is calculated based on the center of gravity of the face area (S8). When the camera 1 is moved, the camera 1 can be moved so that the center of gravity of the face area is located at the center of the image (the intersection of the respective center lines in the X-axis direction and the Y-axis direction).
[0039]
After calculating the movement position of the camera 1, the movement amount of the camera 1, that is, the movement amount of the stage 12 in the driving device 2 is calculated from the current position of the camera 1 and the calculated position of the camera 1. When the movement amount of the stage 12 is calculated, a movement control signal for moving the stage 12 is output to the drive control unit 23. The drive controller 23 moves the stage 12 based on the movement control signal output from the image processor 22. In this way, the imaging position of the camera 1 can be adjusted.
[0040]
As described above, in the imaging position adjustment method according to the present embodiment, the coordinate value of the pixel where the face region exists in the image is added, and the added value is divided by the number of pixels where the face region exists, and the specific position is calculated. Seeking. As described above, since the center of gravity of the position of the face region on the entire surface in the image is obtained, the imaging position can be determined stably even when there is a region change or a detection degree change in the image.
[0041]
Here, a supplementary description will be given of the fact that the imaging position can be determined stably. 4A and 4B show images at the time of determining the imaging position, wherein FIG. 4A is a diagram showing a conventional example, and FIG. 4B is a diagram showing an example of the present embodiment.
[0042]
In the conventional method, when there are two or more face areas, the imaging position is adjusted so that the face area having the largest area is set at the center. Here, for example, as shown in FIG. 4A-1, it is assumed that there are areas E1 and E2 indicating two face areas in the image 30. In this case, if the sizes of the areas E1 and E2 are substantially the same, the direction of the camera frequently changes. For example, when the area E1 of the areas E1 and E2 shown in FIG. 4A is slightly larger, the area E1 indicated by a solid line is located at the center of the image 30 as shown in FIG. Camera orientation is adjusted as follows. Conversely, when the area E2 becomes slightly larger, the camera direction is adjusted so that the area E2 indicated by the solid line is at the center of the image 30, as shown in FIG. For this reason, the direction of the camera frequently changes greatly, and as a result, the imaging position becomes unstable.
[0043]
On the other hand, in the method according to the present embodiment, it is assumed that there are areas E1 and E2 indicating two face areas in the image 30, as shown in FIG. In this case, even if the sizes of the regions E1 and E2 are almost the same and the sizes of the regions E1 and E2 are changed, the direction of the camera is hardly changed as shown in FIG. Can be done. In the present method, since the camera is pointed at the position of the center of gravity of a plurality of regions, even if the size of each region changes, the effect can be reduced. As a result, the imaging position of the camera can be stabilized.
[0044]
In addition, in the imaging position adjustment method according to the present embodiment, the time-consuming division processing is performed only once, so that the processing time can be reduced.
[0045]
Next, a second embodiment of the present invention will be described.
[0046]
The imaging device used in the present embodiment is almost the same as that used in the first embodiment, except for the configuration of the image processing unit 22 and the procedure of the image processing. Hereinafter, the different parts will be mainly described.
[0047]
FIG. 5 is a part of a flowchart illustrating a procedure of the imaging position adjustment method according to the present embodiment. In the imaging position adjustment method according to the present embodiment, first, a flag is set for a pixel in the image 30 where the face image exists in the image 30 through the procedure of steps S1 to S4 of the flow illustrated in FIG. On the other hand, in the image processing unit 22, a centroid memory having the same size and the same number of pixels as the image 30 is formed. Then, voting is performed at a position corresponding to the flagged pixel in the centroid memory. After this process is completed, a vote is performed through the same procedure, and this process is repeated for a predetermined time.
[0048]
This point will be further described with reference to FIG. FIG. 6 is a diagram showing a plurality of images 30A, 30B,. In the image 30A, the regions E1, E2, and E3 are shown at the same positions as in the case shown in FIG. From this image, the pixel in which the face region exists is extracted, and a voting is performed for the barycentric coordinate value corresponding to the pixel in the barycentric memory. The image 30B shows a screen captured after a lapse of a short time after capturing the image 30A, and shows an image slightly different from the image 30A. In this image 30B, as in the case of the image 30A, the pixel in which the face region exists is extracted, and a vote is performed on the barycentric coordinate value corresponding to the pixel in the barycentric memory. Further, the image 30C shows an image captured after a lapse of a short time after capturing the image 30B, and an image slightly different from the image 30B is displayed. In the image 30C, voting for the barycentric coordinate value of the barycentric memory is performed in the same manner.
[0049]
In this way, the pixels in the face area of each of the images 30A, 30B,... Are voted for the barycentric coordinates of the barycentric memory (S11). When the region where the face image exists in each of the images 30A, 30B,... Is voted for in the centroid memory, the time interval for photographing each image is very small, so that the centroid memory is as shown in FIG. 7, for example. In FIG. 7, among the barycentric coordinates in which the voting is performed in the barycentric memory, a portion where many votings are performed is drawn dark, but the barycentric coordinates where the voting is performed are concentrated in a part and the voting is performed. The coordinates of the center of gravity are small, and the others are portions where no voting is performed. By voting from a plurality of images that change in response to time changes to the centroid memory, one centroid coordinate of the centroid memory has the number of obtained images. Here, a large number of votes will be performed in the centroid memory, all of which can be processed with integer values. Therefore, the calculation processing in the center-of-gravity memory can be performed in a short time. Then, by adding the voting in each image for a plurality of images, it is possible to calculate the center of movement of the moving recognition target. In addition, since a plurality of images are added, even when noise occurs, the effect of removing the noise can be exerted.
[0050]
In this manner, when the voting for the barycentric coordinate value is performed, the voting portion performed before the voting is performed is deleted from the centroid memory (S12). Subsequently, the average value of the barycentric coordinate values voted in the barycenter memory is calculated (S13). The average value of the barycentric coordinate values can be obtained by dividing the barycentric coordinate value at which voting is performed by the number of barycentric coordinates at which voting is performed. As described above, by dividing the barycentric coordinate value by the number of barycentric coordinates, it is possible to calculate the average value of the barycenters of all the pixels where the face coordinates existed in a plurality of images captured for a predetermined time. The position corresponding to the center of gravity is set as a specific position in the image, and the movement position of the camera 1 is calculated based on the center of gravity (S14). Specifically, the camera 1 is moved so that, for example, the center of gravity is located at the center of the image (the intersection of the respective center lines in the X-axis direction and the Y-axis direction). Thereafter, in the same manner as in the first embodiment, the movement amount of the camera 1, that is, the movement amount of the stage 12 in the driving device 2 is calculated, and a movement control signal for moving the stage 12 is transmitted to the drive control unit 23. Output. Then, the drive controller 23 moves the stage 12 to adjust the imaging position of the camera 1.
[0051]
As described above, in the imaging position adjustment method according to the present embodiment, the center of gravity, which is a specific position of an image, is obtained in consideration of changes in a plurality of images with time. Therefore, the imaging position is less frequently moved, and a stable image can be obtained. Here, the change of the imaging position (center of gravity) in the case where the time change is considered and the case where the time change is not taken into consideration will be described with reference to FIG. When the imaging position was adjusted by the operator, the imaging position changed relatively largely. On the other hand, when the imaging position is adjusted in consideration of the time change as in the present embodiment, the change in the imaging position is small. As described above, by considering the time change, the change in the imaging position can be reduced.
[0052]
In addition, when the centroids of many images are obtained, the division is performed after adding the coordinates corresponding to all the pixels, so that the division process, which is a time-consuming process, can be completed only once. Therefore, the time required for the arithmetic processing can be reduced.
[0053]
In each of the above embodiments, the calculation of the center of gravity and the movement position of the camera 1 are determined based on whether or not a pixel has a face image. For example, a predetermined condition in these pixels is weighted. You can also. The weighted aspect will be described. In each of the above embodiments, when determining whether or not there is a face image, the face image is determined based on whether or not there is a skin color or a color close to the skin color. If no, there is no face image. At this time, for example, the reference color is weighted irrespective of the presence or absence of the face image, the skin color is weighted heavily, and the weighting is gradually reduced for colors close to but not skin color. be able to.
[0054]
By performing such weighting, the face image to be recognized can be more accurately watched, so that even when there is a change in the face image or a change in the detection degree in the image, the imaging position is determined stably. can do. Here, as the condition for performing the weighting, in addition to the color, luminance and lightness can be used, and further, a mode can be adopted in which the weighting is performed using a condition in which these are combined.
[0055]
The preferred embodiments of the present invention have been described above, but the present invention is not limited to the above embodiments. Although a pixel is used as a region where image processing is performed, a mode in which image processing is performed with a predetermined number, for example, two or four pixels, as one region may be adopted. In the second embodiment, the center of gravity is obtained after adding the position corresponding to the face area. However, the time change of the center of gravity obtained in each image is used, and the average value thereof is set as the specific position. You can also.
[0056]
【The invention's effect】
As described above, according to the present invention, an imaging position adjustment method that can stably determine an imaging position even when there is an area variation or a detection degree variation in an image and does not require complicated arithmetic processing And an imaging position adjustment device can be provided.
[Brief description of the drawings]
FIG. 1 is a block diagram of an imaging apparatus including an imaging position adjustment apparatus according to the present invention.
FIG. 2 is a flowchart illustrating a procedure of an imaging position adjustment method according to the present invention.
3A is a diagram illustrating an image in which a face image is projected, FIG. 3B is a diagram illustrating a face region of the image in FIG. 3A, and FIG. 3C is a diagram illustrating a state in which the face region is captured in pixel units; FIG.
4A and 4B show images at the time of determining an imaging position, wherein FIG. 4A is a diagram showing a conventional example, and FIG. 4B is a diagram showing an example of the present embodiment.
FIG. 5 is a part of a flowchart showing a procedure of an imaging position adjustment method according to a second embodiment.
FIG. 6 is a diagram showing a plurality of images captured at short intervals.
FIG. 7 is a diagram illustrating a center-of-gravity memory in which positions where face images are present are voted from a plurality of images captured at short time intervals.
FIG. 8 is a graph showing a time change of an imaging position when a time change is not taken into consideration;
[Explanation of symbols]
REFERENCE SIGNS LIST 1 camera, 2 drive device, 3 recognition control device, 11 base, 12 stage, 21 A / D conversion unit, 22 image processing unit, 23 drive control unit, 30 (30A, 30B, 30C) ... image, E1 to E3 ... area, F1 to F3 ... face, M ... imaging device, P ... pixel.

Claims (10)

An imaging position adjustment method of extracting a recognition target in an image captured by an imaging unit and adjusting an imaging position of the imaging unit so that the recognition target is present in the image,
Determine a specific position in the image by performing weighting according to the area of the recognition target in the image,
An imaging position adjustment method, wherein an imaging position of the imaging unit is determined based on the specific position. The imaging position adjustment method according to claim 1, wherein, when performing the weighting, the specific position is obtained by performing weighting in consideration of the existence time of the recognition target object in the image. An imaging position adjustment method of extracting a recognition target in an image captured by an imaging unit and adjusting an imaging position of the imaging unit so that the recognition target is present in the image,
Add the coordinate values of the area where the recognition target exists in the image,
The value obtained by adding the coordinate values, the position corresponding to the value obtained by dividing the number of areas where the recognition target exists is determined as a specific position in the image,
An imaging position adjustment method, wherein an imaging position of the imaging unit is determined based on the specific position. The addition of the coordinate values is repeatedly performed a plurality of times for a predetermined time,
When determining the specific position in the image, the value obtained by adding the coordinate values during the predetermined time is divided by the number of times that the area where the recognition target object is present during the predetermined time is added. The method according to claim 3, wherein a position corresponding to the value is determined as a specific position in the image. An imaging position adjustment method of extracting a recognition target in an image captured by an imaging unit and adjusting an imaging position of the imaging unit so that the recognition target is present in the image,
Weighting the coordinate values of the area that satisfies the predetermined condition,
Add the weighted coordinate values of the area satisfying the predetermined condition,
A value obtained by adding the coordinate values is obtained as a specific position in the image, the position corresponding to a value obtained by dividing the value obtained by dividing the number of regions satisfying the predetermined condition,
An imaging position adjustment method, wherein an imaging position of the imaging unit is determined based on the specific position. The weighting of the coordinate values is repeatedly performed a plurality of times during a predetermined time,
When obtaining a specific position in the image, a position corresponding to a value obtained by dividing the value obtained by adding the coordinate values by the number of times of adding the area satisfying the predetermined condition during the predetermined time is defined as the image. The imaging position adjustment method according to claim 5, wherein the imaging position is determined as a specific position in the image. Image processing means for extracting a recognition target in an image picked up by an imaging means, and imaging position control means for performing control for adjusting an imaging position of the imaging means so that the recognition target is present in the image; With
The image processing unit is a coordinate value adding unit that adds a coordinate value of an area where the recognition target exists in the image,
A specific position calculating unit that obtains, as a specific position in the image, a position corresponding to a value obtained by dividing the value obtained by the coordinate value unit by the number of regions where the recognition target exists,
An imaging position determining unit that determines an imaging position of the imaging unit based on the specific position;
An imaging position adjusting device comprising: The coordinate value adding means, when adding the coordinate value, for a predetermined time, calculates the coordinate value of the area where the recognition target is present,
The specific position calculating means, when obtaining a specific position in the image, the value obtained by the coordinate value adding means during the predetermined time, the area where the recognition target object is present during the predetermined time The imaging position adjustment device according to claim 7, wherein a position corresponding to a value obtained by dividing by the number of times of addition is obtained as a specific position in the image. Image processing means for extracting a recognition target in an image picked up by an imaging means, and imaging position control means for performing control for adjusting an imaging position of the imaging means so that the recognition target is present in the image; With
Weighting means for weighting coordinate values of an area satisfying a predetermined condition,
Coordinate value adding means for adding the weighted coordinate values of the area satisfying the predetermined condition,
A specific position calculation unit that obtains, as a specific position in the image, a position where a value obtained by dividing the value obtained by the coordinate value adding unit by the number of regions that satisfy the predetermined condition is used as a specific position; As an imaging position determining means for determining an imaging position of the imaging means,
An imaging position adjusting device comprising: The weighting means, when performing the weighting, for a predetermined time, weights the coordinate value of an area that satisfies a predetermined condition,
The specific position calculation means calculates a value obtained by dividing the value obtained by adding the coordinate values by the number of times of adding the area satisfying the predetermined condition during the predetermined time when obtaining the specific position in the image. The imaging position adjustment device according to claim 9, wherein a position corresponding to the position is determined as a specific position in the image.

Images