JP2002091686A - Method and device for detecting position, visual feature detecting method, pointing system and monitoring system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide position detecting method/device for precisely detecting the two-dimensional position of an object with a small constitution without the need of conventional scanning mechanism and driving means. SOLUTION: Omniazimuth photographing units 1 which can photograph an omniazimuth in a peripheral direction at a time are individually arranged near the four corner parts of a panel 51 with which the tip of an instructing pen 52 that a user operates is brought in contact. A main control part 53a detects the position of the instructing pen 52 on the basis of the picture of the omniaziumth including the instructing pen 52 which are acquired with the respective omniazimuth photographing units 1. A moving track as the secular position fluctuation of the instructing pen 52 is displayed on a display part 53c.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method and an apparatus for detecting the position of an object using a plurality of image acquisition means for acquiring images in the circumferential direction, a method for detecting the visual characteristics of an object, and a method for detecting the visual characteristics of the object. The present invention relates to a pointing system and a monitoring system used.

[0002]

2. Description of the Related Art As a method for detecting a two-dimensional position of an object in a plane, a method using a light beam such as a laser beam has been conventionally known. The method using a light beam is a method often used particularly with a pointing device or the like, and detects a reflected light from an object (pointing pen) by performing angular scanning with a laser beam, and detects an object (pointing pen) based on the reflected light. )
Method of detecting the position of the laser beam, or conversely, the laser beam is angularly scanned in a plane whose periphery is covered with a retroreflector, and reflected light cannot be obtained due to the presence of an object (pointing pen) There is a method of detecting the position of an object (pointing pen) by using a method.

As another method for detecting the two-dimensional position of an object, a method using image capturing means such as a camera is also known. In this method, one or a plurality of cameras are arranged, an image including an object to be detected is acquired, the object is identified based on the characteristics (color, shape, etc.) of the object by analyzing the acquired image, and the position is identified. To detect.

[0004]

The above-described method using a light beam has a problem that a scanning mechanism for scanning a light beam such as a laser beam is required. Further, attenuation of the reflected light is also a problem, and there is a problem that it is difficult to detect the position of a distant object.

On the other hand, in the method using the conventional image photographing means, since the field of view of a single image photographing means (camera) is narrow, an extremely large number of images are required to detect the position of an object over a wide range. There is a problem that a photographing means (camera) must be provided. In addition, it is possible to reduce the number of cameras installed by rotating the camera, but in this case, a camera driving means is required, and in addition, an omnidirectional or wide-angle image in the circumferential direction is instantaneously obtained. Cannot be obtained, there is a problem that a time lag occurs in the detection result.

The present invention has been made in view of the above circumstances, and uses a plurality of image acquisition means capable of photographing all directions in the circumferential direction or a wide angle at a time, thereby providing a conventional scanning mechanism, driving means, and the like. A position detection method and a position detection device capable of accurately detecting a two-dimensional position of an object with a low-cost, small-sized configuration that is unnecessary, and a visual feature detection method capable of accurately detecting a visual characteristic of an object, and It is an object to provide a pointing system and a monitoring system as those applications.

[0007]

According to a first aspect of the present invention, there is provided a method for detecting a position of one or a plurality of objects in a plane, wherein a light incident portion is provided at a position corresponding to the plane. A plurality of image acquisition means acquire an image in the circumferential direction including the object, and the position of the object is detected based on the acquired images.

According to the position detecting method of the present invention, a plurality of image obtaining means obtain circumferential images including a detection target object, and analyze the obtained images to detect the position of the object. I do. Therefore, since all directions, that is, a 360-degree range or a wide-angle range are photographed at a time, a mechanism for driving the image acquisition unit is unnecessary, and even with a small number of image acquisition units, the features (color, shape, , Etc.)
The position of an object is accurately detected.

According to a second aspect of the present invention, there is provided a position detecting method according to the first aspect.
In each of the image acquisition units, the position of each of the image acquisition units is fixed, and the first image in the circumferential direction not including the object acquired by each of the image acquisition units is acquired by each of the image acquisition units. A difference image from the second circumferential image including the object is obtained, and the position of the object is detected according to the obtained difference image.

According to a second aspect of the present invention, a difference image between a reference image not including an object and an image including the object is obtained, and the position of the object is detected from the difference image. Therefore, an object existing in an arbitrary direction can be easily identified by the background difference, and the azimuth of the object in each image acquisition unit can be easily obtained.

According to a third aspect of the present invention, there is provided a position detecting method according to the first aspect.
In the second aspect, the image in the circumferential direction is temporally acquired by each of the image acquiring means, and the temporal position of the object is detected.

According to a third aspect of the present invention, an image including an object is acquired with time, and the moving object is tracked. At this time, to obtain an omnidirectional image including the object,
There is no time lag in tracking the object.

According to a fourth aspect of the present invention, there is provided a position detecting method according to the first aspect.
In any one of ~, when the cross-sectional shape of the object by the plane is substantially circular, the error of the azimuth angle with respect to the object in each of the image acquisition means, and the error is obtained by each of the image acquisition means. The position of the object is detected in consideration of an apparent size error of the image of the object.

In the position detecting method according to the fourth aspect, the position of the object is detected in consideration of the error of the azimuth angle with respect to the object and the error of the apparent size of the object in the image as the detection error. Therefore, the position of an object having a substantially circular cross section can be accurately detected.

According to a fifth aspect of the present invention, in the device for detecting the position of one or a plurality of objects in a plane, the position detection apparatus is provided at a position corresponding to the plane, and detects a circumferential image including the object. It is characterized by comprising a plurality of image acquiring means for acquiring, and a detecting means for detecting the position of the object based on the plurality of images acquired by each of the image acquiring means.

In the position detecting device according to a fifth aspect, a plurality of images in the circumferential direction including the object to be detected are obtained by each of the plurality of image obtaining means, and the obtained images are analyzed by the detecting means to analyze the object. Detect the position. Therefore, since all directions, that is, a 360-degree range or a wide-angle range are photographed at a time, a mechanism for driving the image acquisition unit is unnecessary, and even with a small number of image acquisition units, the features (color, shape, ), And accurately detects the position of the object.

According to a sixth aspect of the present invention, there is provided a position detecting device.
Wherein each of the image acquisition means includes a light reflector having a rotationally symmetric shape, an imager provided at a position opposed to the light reflector, and a light incident range on the light reflector. And a regulating member that performs the control.

In the position detecting device according to the sixth aspect, necessary light which is not regulated by the regulating member is reflected by the light reflector having a rotationally symmetric shape, and then is incident on the image pickup device to include the object. A circumferential image is obtained. Since unnecessary light is regulated by the regulating member, an omnidirectional or wide-angle image on an unnecessary plane cannot be obtained, and an image including an object can be reliably obtained.

According to a seventh aspect of the present invention, in the method for detecting visual characteristics of one or a plurality of objects in a plane, a plurality of light incident portions are provided at positions corresponding to the plane. Each of the image acquiring means acquires an image in the circumferential direction including the object, and detects a visual characteristic of the object based on the acquired plurality of images.

According to the visual feature detecting method of the present invention, each of the plurality of image acquiring means acquires a circumferential image including the object to be detected, and analyzes the acquired images to visually recognize the object. Detect characteristic features (shape, color, etc.). Therefore, since all directions, that is, a 360-degree range or a wide-angle range are photographed at a time, a mechanism for driving the image acquisition unit is unnecessary, and even with a small number of image acquisition units, the visual characteristics of the object can be accurately determined. Detect well.

A pointing system according to claim 8, wherein in the pointing system for detecting a position pointed by the pointer, the contacted body contacted by the pointer and the pointer contacted by the contacted body. A plurality of image acquisition means for acquiring images in the circumferential direction including, and a detection means for detecting a position indicated by the pointer based on the plurality of images acquired by each of the image acquisition means, I do.

According to the pointing system of the present invention, a plurality of images in the circumferential direction including the pointing object contacted by the contacted object are acquired by the plurality of image acquiring means, and the detecting means is provided to the detecting means based on the acquired plurality of images. To detect the position indicated by the pointer. Therefore, position detection can be performed with a simple configuration without providing a scanning mechanism and a driving unit.

According to a ninth aspect of the present invention, in the system for monitoring entry of an object into a predetermined plane area, a plurality of image acquisition means for acquiring images in a circumferential direction in the plane area; Detecting means for detecting intrusion of the object based on a plurality of images obtained by each of the detecting means.

According to a ninth aspect of the present invention, a plurality of image acquiring means acquires images in a circumferential direction in a predetermined plane area, and the detecting means detects intrusion of an object based on the acquired plurality of images. I do. Therefore, the intrusion of an object can be detected with a simple configuration without providing a scanning mechanism and a driving unit.

[0025]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be specifically described with reference to the drawings showing the embodiments. First, the configuration of an omnidirectional photographing device used in the present invention that can photograph in all directions, that is, 360 degrees in the circumferential direction at a time, will be described.

FIG. 1 is a diagram showing the configuration of a first example of the omnidirectional photographing device 1. As shown in FIG. In FIG. 1, reference numeral 10 denotes a transparent cylinder made of glass or plastic material. One end of the cylindrical body 10 has a hemispherical convex mirror 11 having a rotationally symmetric shape.
Is fixed to a supporting body 12 for mounting and supporting. The convex mirror 11 is made of brass or the like, and its surface is mirror-finished. The outer peripheral surface of the cylindrical body 10 is covered with a light shielding material 13 except for the light incident portion 10a. The other end side of the cylindrical body 10 is provided with a window hole 14 for light transmission.
Is connected to a camera 16 via a connecting member 15 having

The camera 16 includes a hollow cylindrical housing 17 having one end connected to the connecting member 15, a lens barrel 18 screwed to the inner peripheral surface of the housing 17, and a lens 19 fixed to the lens barrel 18.
And a CCD element 20 attached to the other end of the housing 17. The camera 1 is located at a position facing the top of the convex mirror 11.
The cylindrical body 10 and the housing 17 are connected via the connecting member 15 in a mode in which 6 is arranged. The optical axis of the camera 16 is
The axis coincides with the axis of the convex mirror 11 (the rotation axis of the rotationally symmetric body).
The light incident through the light incident portion 10a and reflected by the convex mirror 11 is focused by the lens 19 and is imaged on the CCD element 20. By adjusting the screwing position of the lens barrel 18 with respect to the housing 17, the position of the lens 19 in the optical axis direction can be changed, and the optical distance between the lens 19 and the CCD element 20 can be adjusted so that focusing can be performed. Has become.

With such a configuration, the omnidirectional photographing device 1
Reflects light from a predetermined height at the light incident portion 10a downward by the convex mirror 11 and forms an image on the CCD element 20 through the window hole 14 and the lens 19, with the optical axis of the camera 16 as the center. Obtain images in all directions (360 °).

In the first example described above, the shape of the convex mirror 11 is hemispherical, but it may be conical or parabolic, and the shape may be a quadratic curved surface. preferable.

FIG. 2 is a diagram showing a configuration of a second example of the omnidirectional photographing device 1. As shown in FIG. The omnidirectional imaging device 1 of the second example uses a concave mirror 21 instead of the convex mirror 11 of the first example. Other configurations are the same as those of the first example described above, and the same portions are denoted by the same reference numerals and description thereof will be omitted.

FIG. 3 is a diagram showing the configuration of a third example of the omnidirectional photographing device 1. In FIG. 3, reference numeral 30 denotes a hollow cylindrical housing made of a light-shielding material, and a support 32 for attaching and supporting a spherical mirror 31 having a true spherical shape is connected to one end of the housing 30. The support 32 is made of glass or plastic material, and extends in the axial direction of the omnidirectional photographing device 1, and the extending portion is a guide portion 32 for introducing external light.
a. The spherical mirror 31 is formed by processing brass or the like, and its surface is mirror-finished. The other end of the housing 30 is a camera 33.

The camera 33 has a lens barrel 34 screwed to the inner peripheral surface of the housing 30, a lens 35 fixed to the lens barrel 34, and a CCD element 36 attached to the other end of the housing 30. . Light from a predetermined height at the guide portion 32a enters through the guide portion 32a and is reflected by the spherical mirror 31, and then is converged by the lens 35 and imaged on the CCD element 36, thereby forming the optical axis of the camera 33. Is obtained in all directions (360 °). This guide part 32a
Prevents light from an oblique direction from being incident on the omnidirectional imaging device 1. Also, as in the first example, the housing 3 of the lens barrel 34
By adjusting the screw position with respect to the lens 35,
Can be changed in the optical axis direction, and the optical distance between the lens 35 and the CCD element 36 can be adjusted to perform focusing.

FIG. 4 is a diagram showing a configuration of a fourth example of the omnidirectional photographing device 1. As shown in FIG. 4, the same parts as those in FIG. 1 are denoted by the same reference numerals, and the description thereof will be omitted. In the omnidirectional photographing device 1 of the fourth example, the distal end side is projected by a predetermined length toward the camera 16 on the extension of the axis of the convex mirror 11, and the linear body 4
1 is fitted and fixed to the convex mirror 11.

The reflected light from the inner surface of the cylindrical body 10 is condensed by the convex mirror 11 on the lens 19 of the camera 16 and forms an image on the CCD device 20, which may lower the accuracy of the acquired image. Therefore, the linear member 41 is provided to prevent this, that is, to prevent unnecessary reflected light from being guided to the camera 16 side. Since the light that is reflected by the inner surface of the cylindrical body 10 and reaches the convex mirror 11 always crosses the extension of the axis of the convex mirror 11 before being reflected by the inner surface, a linear shape is formed at that position as in the fourth example. By providing the body 41, all of the light that is internally reflected and reaches the convex mirror 11 is
It can be blocked with 1. Therefore, the accuracy of the acquired image can be improved.

The linear member 41 as shown in the fourth example
It is needless to say that the omnidirectional photographing device 1 can be configured to be attached to the concave mirror 21 of the second example and the spherical mirror 31 of the third example.

Next, a method for detecting the position of an object using the omnidirectional photographing device 1 having such a configuration will be described. According to the present invention, an object is identified and its position is detected using N (N ≧ 2) omnidirectional imaging devices 1 without using the feature (color, shape, etc.) of the object. FIG. 5 is a flowchart showing the procedure of the detection method.

(Step S1: Acquisition of Omnidirectional Image) An omnidirectional image including an object to be detected is acquired by N omnidirectional photographers 1 fixed at substantially the same height in a certain environment.

(Step S2: Acquisition of Difference Image) An omnidirectional image (background image) previously acquired by each omnidirectional photographing device 1 in a state where no object to be detected exists, and an omnidirectional image acquired in step S1 Is obtained.

(Step S3: Detection of Object Azimuth) Based on the difference image obtained in step S2, an object existing in an arbitrary direction is identified, and the position of the object with respect to each omnidirectional camera 1 is determined from the position on the image. Detect azimuth.

(Step S4: Estimation of In-Plane Position of Object (Two-Dimensional Position)) If the installation position and orientation of each omnidirectional photographing device 1 are known, stereo vision (triangle) is performed from the azimuth detected in step S3. Surveying) to estimate the position (two-dimensional position) of the object in the plane.

If it is known that only one object exists in the detection target area, the position of the object can be estimated using at least two omnidirectional imaging devices 1. On the other hand, if there are multiple objects in the detection target area,
Alternatively, when it is unknown how many objects are present, the positions of those objects can be estimated using at least three omnidirectional imaging devices 1. When a larger number of omnidirectional imaging devices 1 are used than the minimum, the position estimation accuracy is improved.

When it is known that there is only one object, as shown in FIG. 6A, the azimuth information of the two omnidirectional imaging devices 1a and 1b (solid line in the figure) is used. The position of the object can be estimated as their intersection (A in the figure). However, when there are a plurality of objects, as shown in FIG. 6B, from the azimuth information (solid lines in the figure) of the two omnidirectional imaging devices 1a and 1b,
Since it is considered that the object exists at four positions (A to D in the figure), the accurate position of the object cannot be estimated. Therefore, by adding azimuth information (broken line in the figure) in one omnidirectional photographing device 1c, positions where no object actually exists (C and D in the diagram) are deleted, and the correct position of the object is corrected. (A and B in the figure) can be estimated.

Incidentally, in such a position estimation, a problem of an observation error occurs. For example, at points A and B shown in FIG. 6B, three omnidirectional imaging devices 1a and 1
Although the three azimuths b and 1c intersect, actually, the obtained azimuths include an error, and there is a high possibility that they do not intersect at one point. Further, when the distance between the object and the omnidirectional photographing device is relatively short, the object is identified with a certain width, which also causes an observation error. Further, when an object exists near a straight line (base line) connecting the two omnidirectional photographing devices, the difference in azimuth angle in each omnidirectional photographing device almost disappears, and the intersection becomes closer to the point at infinity. Poor precision of vision. Therefore, it is necessary to estimate the position of the object in consideration of the observation error as described above.

In the present invention, in order to reduce such an observation error, the position of the object is estimated by using N (N ≧ 3) omnidirectional imaging devices 1. FIG. 7 shows a procedure for estimating the position of an object,
That is, it is a flowchart showing a subroutine of step S4.

First, the position of an object is determined by stereo vision (binocular stereo vision) based on the azimuth of the object obtained by any two of the N omnidirectional photographers 1 (two-lens stereovision). (See A to D in FIG. 6B) (Step S4)
1). The obtained position of the object is referred to as an estimated position in binocular stereo vision.

The estimated position in (k-1) stereoscopic vision is obtained by another omnidirectional camera 1 different from the (k-1) omnidirectional cameras 1 from which the position is obtained. To check whether the azimuth angle of the object to be checked matches k =
Repeat from 3 to k = N (Steps S42-3 to S4
2-N). For the azimuth angles of all N azimuth photographers 1,
The number of overlaps (the number of overlaps M) of N hexagonal regions as described later is obtained (step S43). The above-described processing is performed for all pairs of arbitrary two omnidirectional imaging devices 1 in the N omnidirectional imaging devices 1 (step S44: NO).

Here, a specific method of considering the observation error will be described. In the following example, the cross section of the object in the position detection target plane is circular. In this case, as the observation error, a detection error α at the right end and the left end of the object, which is an error of azimuth angle detection, and an error β of a diameter of a circle representing the object are considered.
When the error is modeled in this way, the center position of the circle representing the object can be estimated as a hexagonal region indicating these error ranges, as described below.

First, it is assumed that an object C having a diameter d is obtained from the omnidirectional photographing device 1 as shown in FIG. Here, consider a circle C ₋ having a diameter of d−β. The condition that capture the rightmost-leftmost circle C _{- -} right end and circle from the left end in the range of angles ± alpha C of the object C captured by the omnidirectional imaging apparatus 1 range center is present in the FIG. 8 (a ) Are indicated by the hatched areas. Here, the straight lines n and n 'and the straight lines m and m' are parallel to each other, and FIG. 8A shows only the upper half region. Similarly, when a circle C ₊ having a diameter of d + β is considered as shown in FIG.
This is represented by the hatched area in FIG. here,
The straight lines n and n ″ and the straight lines m and m ″ are parallel to each other.
(B) shows only the upper half area. Therefore, the center of the circle whose diameter is in the range of d−β to d + β is shown in FIG.
FIG. 8 includes the vertices (x marks) of the regions shown in FIGS.
It exists in a hexagonal area as shown in FIG. The hexagonal area represents the estimated position of the object when the measurement errors α and β are considered.

When estimating the position of an object in consideration of such observation errors, whether or not the object can be identified by the azimuths of the plurality of omnidirectional imaging devices 1 depends on whether the hexagonal areas overlap. It can be checked by whether or not.

For example, the estimated positions of the object in the binocular stereo vision by the two omnidirectional imaging devices 1 (1a, 1b) obtained in step S41 are two 6-position images as shown in FIG.
It is represented by the overlapping portion of the rectangular area. Similarly, k pieces (k = 3) obtained in steps S52-3 to S52-N described above.
The estimated position of the object in the k-eye stereo vision by the omnidirectional imaging device 1 of (N) to (N) is represented by an overlapping portion of k hexagonal regions.

As a result of examining the azimuth angles of all the N omnidirectional photographing devices 1, if the number of overlaps M in such a hexagonal area is large, it is highly possible that an object exists there. Therefore, if it is known that there is only one object existing in the detection target area (step S45: YES), the object having the largest number of overlaps M is detected as the position of the object (step S46). On the other hand, when there are a plurality of objects in the detection target area, or when it is unknown how many objects exist in the detection target area (step S45: N
O), the object whose overlap number M is larger than a predetermined threshold T is detected as the position of the object (step S47). On this occasion,
If the number of existing objects is known, the object having the larger number of overlaps M is preferentially adopted.

By the above-described processing, the position of the erroneous object detected in the binocular stereo vision (C, C in FIG. 6B)
D) is removed, and only the correct position of the object (A and B in FIG. 6B) can be accurately detected.

If the object and the omnidirectional photographing device 1 are in a special positional relationship, there is a possibility that an erroneous detection will occur. For example, as shown in FIG. 10, when three omnidirectional imaging devices 1a, 1b, and 1c are used, it is considered that an object also exists in D in addition to A, B, and C where an object actually exists. . Such an error in detecting the position D occurs when the azimuth used for estimating the position of the object is also used for estimating the position of another object. In the example of FIG. 10, one azimuth angle by the omnidirectional imaging device 1a is used for estimation of the positions B and D, and one azimuth angle by the omnidirectional imaging device 1b overlaps the estimation of the positions A and D. Azimuth angle by the omnidirectional imaging device 1c
Used in the estimation of Therefore, if the azimuths used for estimating the position of a certain object are all used twice or more, it is highly likely that there is an error, and the detected position is removed or It is desirable to verify the detection position based on the feature (color, shape, etc.) of the object.

(Step S5: Storing (Displaying) Position Information of Detected Object) Information on the position of one or more objects detected in step S4 is stored in a predetermined medium or in real time. indicate.

(Step S6: Judgment of end instruction) When an instruction to end the process of detecting the position of the object is given (S6:
6: YES), the process ends, and when the instruction is not given, for example, when the position of the object is detected over time such as when tracking the object (S6: NO), the processing from step S1 is repeated. Repeat the process.

In the above-described method of estimating the position of an object by N-eye stereo vision, processing for examining the overlap of hexagons is required, and the amount of calculation is large, which is not suitable for real-time processing. Conceivable. Therefore, a slightly simplified method will be described below.

In the binocular stereo vision stage, a circle having a predetermined diameter representing an object centered on an intersection of straight lines representing the azimuth angles of the objects detected by any two omnidirectional imaging devices is set as the position of the object. . FIG. 11 shows such an estimated position of the object. In the example of FIG. 11, three omnidirectional imaging devices 1a, 1b,
The three estimated positions (circles) of the object in the binocular stereo vision obtained by the pair of 1c are indicated by solid lines. Also,
In N-eye stereo vision, it is determined whether or not an object is detected by the Nth omnidirectional imaging device based on whether or not those circles overlap. If they overlap, the circle having the predetermined diameter centered on the center of gravity is set as the position of the object.
Specifically, in the example of FIG. 11, a circle centered on the barycentric position of the center of the three circles (dashed circle in FIG. 11) is set as the estimated position of the object in three-lens stereo vision.

The observation errors α and β in the simplified method are considered as follows. As described above, in the vicinity of a straight line connecting two omnidirectional photographing devices, the position estimation accuracy of an object by stereo vision deteriorates. Therefore, this simplified method has the following problems. For example, FIG.
As shown in FIGS. 2 (a) to 2 (c), at the stage of binocular stereo vision, an intersection is not obtained due to an azimuth angle observation error. The circle to represent may not be obtained. In addition, as shown in FIG. 13, in the stage of N-eye stereo vision, circles obtained by (N−1) -eye stereo vision do not overlap due to an azimuth observation error, and the position of an object may not be estimated. is there.

Therefore, such a problem is solved by the following processing. (1) When Two Straight Lines Do Not Have an Intersection Two omnidirectional imaging devices 1 as shown in FIGS.
The angles θ _a and θ _b between the straight line n connecting the lines a and b and the azimuths of the objects detected by the omnidirectional imaging devices 1 a and 1 _b are both α.
If it is less, considered as the object is on the straight line n, both the omnidirectional imaging unit 1a, 1b the ratio of the distance of each of between the object d _a: d _b is omnidirectional imaging device 1a, 1b So that the ratio of the reciprocal of the apparent size of the object seen from
Correct the position of the object on the straight line n. FIG. 12 (c)
, One of these angles θ _a and θ _b (FIG. 1)
In the example of FIG. 2 (c), when only θ _b ) is equal to or smaller than α, it is considered that the object is on the straight line representing the other azimuth angle (θ _a ), and the omnidirectional imaging devices 1a and 1b and the object the ratio d _a distance each between: d _b is to be closest to the ratio of the reciprocal of the size of the object apparent, to correct the theta _b only Δθ (Δθ ≦ α).

(2) When Two Circles Do Not Overlap As shown in FIG. 13, at least one of the circles
If the azimuth angle can be corrected to α or less (in the example of FIG. 13, the azimuth angle by the omnidirectional imaging device 1b is corrected by Δθ), if the azimuth angle can be overlaid on the other circle, those 2
The individual circles are considered to be overlapping.

The observation error β is considered as follows. From the distance between the estimated position of the object and the omnidirectional camera, the size (width) of the object reflected on the omnidirectional camera can be estimated. If the estimated size differs from the actually observed size by β or more, it is regarded as erroneous detection, and it is determined that no object exists at that position.

Next, embodiments of some systems to which such an omnidirectional photographing device is applied will be described.

(First Embodiment) First, an example of a pointing system will be described. FIG. 14 is a diagram showing a configuration of the pointing system according to the first embodiment of the present invention. In FIG. 14, reference numeral 51 denotes a flat panel as a contacted body. The tip of the pointing pen 52 as an indicator operated by the user is brought into contact with the panel 51. Near the four corners of the panel 51, the first
One omnidirectional photographing device 1 of any of the examples to the fourth example is provided. Note that each of these four omnidirectional imaging devices 1 is provided in such a manner that the light incident portion 10a (or the guide portion 32a) includes a plane forming the upper surface of the panel 51, and the tip of the pointing pen 52 Obtain an omnidirectional image including.

Each omnidirectional photographing device 1 is connected to a position detector 53. The position detector 53 includes a main control unit 53a, a storage unit 53b, a display unit 53c, a recording unit 53d, and a ROM 53.
e, a RAM 53f, an input interface 53g, and the like.

The main control unit 53a is specifically constituted by a CPU, controls each of the above-mentioned hardware components in the position detector 53, and controls an object (instruction) according to a computer program stored in the ROM 53e. Various software functions such as position estimation of the pen 52) are executed.

The storage unit 53b stores the detected pointing pen 52
Is stored. The display unit 53c displays the detected position data of the pointing pen 52. In this display mode, the position information (two-dimensional coordinates) itself may be displayed, or the movement locus of the position of the pointing pen 52 may be displayed. The recording unit 53d includes a storage unit 53
Read the position data from b and print out on paper.

The ROM 53e stores various software programs necessary for operations such as the position estimation of the pointing pen 52 in advance. The RAM 53f is configured by an SRAM, a flash memory, or the like, and stores temporary data generated when executing software. The input interface 53g controls input of an acquired image from each omnidirectional imaging device 1.

Next, the operation will be described. The pointing pen 52 is moved by a user's operation in such a manner that its tip is in contact with the panel 51. An omnidirectional image including the pointing pen 52 is acquired by the four omnidirectional imaging devices 1 provided at four corners of the panel 51, and the acquired image is input to the position detector 53. In the position detector 53, each omnidirectional imaging device 1
The position of the pointing pen 52 is detected based on the image obtained in step (1) and according to the above-described algorithm (four-lens stereo vision). Then, for example, the position of the pointing pen 52 is displayed on the display unit 53c, and the temporal change in the position is displayed on the display unit 53c as a movement locus.

(Second Embodiment) Another example of the pointing system will be described. FIG. 15 shows a second embodiment of the present invention.
FIG. 1 is a diagram illustrating a configuration of a pointing system according to an embodiment. In FIG. 15, reference numeral 54 denotes a flat screen. The tip of the pointing pen 52 operated by the user is brought into contact with the screen 54. In the vicinity of the four corners of the screen 54, one omnidirectional photographing device 1 of any of the above-described first to fourth examples is provided. Note that each of these four omnidirectional imaging devices 1 has its light incident portion 10
a (or the guide portion 32 a) is provided in a mode including the surface of the screen 54, and acquires an omnidirectional image including the tip of the pointing pen 52.

A projector 55 is provided below the screen 54. The projector 55 is connected to a computer 56 and projects predetermined video data supplied from the computer 56 onto a screen 54. Each omnidirectional camera 1 is connected to the computer 56. The computer 56 has the same function as the above-described position detector 53. The same image as the projected image on the screen 54 is displayed on the display screen 56a, and the movement locus of the detected pen 52 is detected. Is also displayed.

Next, the operation will be described. Projector 55
, A predetermined image is displayed on the screen 54.
At this time, the same video is displayed on the display screen 56a of the computer 56. In such a state, the user operates and moves the pointing pen 52 in such a manner that the tip touches the screen 54. An omnidirectional image including the pointing pen 52 is obtained by the four omnidirectional imaging devices 1 provided at four corners of the screen 54, and the obtained image is input to the computer 56. In the computer 56, based on the image acquired by each omnidirectional camera 1,
The position of the pointing pen 52 is detected according to the algorithm of stereoscopic vision. Then, the movement trajectory of the pointing pen 52 is displayed on the display screen 56a in a manner overlapping the predetermined image. Therefore, it is possible to conveniently display a desired image that the user wants to add to the existing video image.

(Third Embodiment) An example of a monitoring system will be described. FIG. 16 is a diagram illustrating a configuration of a monitoring system according to the third embodiment of the present invention. In FIG. 16, reference numeral 57 denotes a wall of a museum, and a painting 58
Is hung. The four omnidirectional imaging devices 1 of any of the first to fourth examples described above are provided on the wall surface 57 at a position surrounding the painting 58. Note that each of these four omnidirectional imaging devices 1 is provided in such a manner that its light incident portion 10a (or guide portion 32a) is positioned at the height of the surface of the painting 58, and Obtain an omnidirectional image in a plane.

Each omnidirectional photographing device 1 is connected to a position detector 59. The position detector 59 includes a main control unit 59a, a display unit 59b, a ROM 59c, a RAM 59d, an input interface 59e, an alarm 59f, and the like.

The main control unit 59a is specifically constituted by a CPU, controls each of the above-described hardware units in the position detector 59, and detects a suspicious object in accordance with a computer program stored in the ROM 59c. Perform various software functions such as position estimation. The display unit 59b displays an image acquired by each omnidirectional camera 1 as necessary. The ROM 59c stores in advance various software programs required for operations such as position estimation of an object (suspicious object). The RAM 59d is configured by an SRAM, a flash memory, or the like, and stores temporary data generated when executing software. The input interface 59e controls the input of the acquired image from each omnidirectional imaging device 1. The alarm 59f issues an alarm when it is detected that a suspicious object exists at a predetermined position.

Next, the operation will be described. Images in all directions including a predetermined plane are constantly acquired by the four omnidirectional imaging devices 1 provided on the wall surface 57, and the acquired images are input to the position detector 59. In the position detector 59, detection of the presence of an object (suspicious object) and detection of the position in the presence of an object (suspicious object) are performed in accordance with the above-described four-lens stereoscopic vision algorithm based on the image acquired by each omnidirectional imaging device 1. Will be When an object is detected at a predetermined position, the object is regarded as a suspicious object, and an alarm is issued from the alarm 59f. Therefore, not only can the theft of the painting 58 be prevented, but also the painting 58 can be prevented.
Contact can also be prevented.

(Fourth Embodiment) An example of a security system will be described. FIG. 17 is a diagram showing a configuration of a maintenance system according to the fourth embodiment of the present invention. In FIG. 17, reference numeral 60 denotes an entrance of a space having a high radiation environment inside, or a space in which a dangerous machine tool is operating. At the position surrounding the entrance 60, the first
Four omnidirectional imaging devices 1 of any of the examples to the fourth examples are provided. In addition, each of these four omnidirectional imaging devices 1 is provided in such a manner that the light incident portion 10a (or the guide portion 32a) is aligned with the position of the entrance 60, and the omnidirectional photographing device 1 is arranged on the plane at that position. Acquire the image of the bearing. Each omnidirectional imaging device 1 is connected to a position detector 59 similar to that of the third embodiment.

Next, the operation will be described. Entrance 60
An omnidirectional image including the surface is constantly acquired by four omnidirectional imaging devices 1 provided in the vicinity, and the acquired image is input to the position detector 59. In the position detector 59,
Based on the image acquired by each omnidirectional camera 1, detection of the presence of an object (suspicious object) and detection of the position when the object is present are performed in accordance with the algorithm of the four-eye stereo vision as described above. When an object is detected at the position of the entrance 60,
It is regarded as a suspicious object, and an alarm is issued from the alarm 59f. Therefore, if the dangerous entrance 60 is attempted to be accessed by mistake, a warning is issued and the user is notified, so that danger can be prevented beforehand and safety can be ensured.

(Fifth Embodiment) An example of an object shape measuring system will be described. FIG. 18 is a diagram showing a configuration of a shape measuring system according to the fifth embodiment of the present invention. In FIG. 18, four omnidirectional imaging devices 1 of any of the above-described first to fourth examples are provided so as to surround a predetermined two-dimensional space. Note that all of these four omnidirectional imaging devices 1 are provided in such a manner that the light incident portion 10a (or the guide portion 32a) is positioned so as to include the two-dimensional space. Obtain an omnidirectional image at. The object 61 to be detected moves while passing through the two-dimensional space. Each omnidirectional camera 1 is connected to a computer 56, and the acquired omnidirectional image is input to the computer 56.

In the case where the moving speed of the object 61 is known by detecting the position of the object 61 over time based on the images acquired by the four omnidirectional imaging devices 1 in the computer 56 , The surface shape of the object 61 can be measured.

(Sixth Embodiment) An example of a performance system will be described. FIG. 19 is a diagram showing a configuration of a performance system according to the sixth embodiment of the present invention. In FIG. 19, reference numeral 62 denotes a floor surface, and the omnidirectional imaging device 1 of any of the first to fourth examples described above is provided at each of the four corners of the floor surface 62. Dancer 63 is performing tap dance. Note that each of these four omnidirectional imaging devices 1 has its light incident portion 10a (or guide portion 32a).
Is positioned at the height of the foot of the dancer 63, and obtains an omnidirectional image on a plane at that height. Each omnidirectional camera 1 is connected to a computer 56, and the acquired omnidirectional image is input to the computer 56.

In this way, the movement of the foot of the dancer 63 is detected by analyzing the images acquired over time by each omnidirectional photographing device 1. And dancer 6
It is possible to play rhythm music by recognizing the rhythm from the subtle movement of the third foot.

(Seventh Embodiment) An example of an object detection system will be described. FIG. 20 is a diagram showing the configuration of the object detection system according to the seventh embodiment of the present invention. Two omnidirectional imaging devices 1 of any of the above-described first to fourth examples are provided in two spaces, and each omnidirectional imaging device 1 has a position detector 59 similar to that of the third embodiment. The connected omnidirectional images are connected to the position detector 59.

When a predetermined line is set as an object detection line, a conventional detection system using infrared rays detects a person or an object only when the line arrives. In contrast, in the object detection system of the present invention shown in FIG. 20, when a line (shown by a broken line) connecting both omnidirectional imaging devices 1 and 1 is an object detection line, a person or an object approaches the line. When going, it is possible to detect a person or an object in advance with the trouble of the line.

(Eighth Embodiment) In each of the embodiments described above, the case where the position of one object is basically detected has been described. In the eighth embodiment, a system for detecting the position of one or a plurality of objects over time, in other words, the number of objects whose number is unknown is described with reference to a case where a detection target is a human. .

FIG. 21 is a diagram showing a configuration of a human tracking system according to the eighth embodiment of the present invention. The first mentioned above
Any of the four omnidirectional imaging devices 1 of the examples to the fourth example
It is provided on the support base 65 at a height of about 1 m from the floor in the room 4, and acquires an omnidirectional image at that height position in the room 64. These four omnidirectional imaging devices 1 are connected to the same position detector 53 as in the first embodiment. In this system, the positions and orientations of the four omnidirectional imaging devices 1 are measured before starting the tracking process.

Next, the operation will be described. Omnidirectional images at a height of 1 m are acquired by four omnidirectional imaging devices 1 provided in the room 64, and the acquired images are input to the position detector 53. In the position detector 53,
Based on the image acquired by each omnidirectional camera 1, it is detected whether or not a human is present, and when the presence of a human is recognized, the position is detected. Since the omnidirectional imaging device 1 is fixed in a predetermined environment, a person present in an arbitrary direction can be relatively easily recognized by the background difference, and the azimuth angle of the person can be obtained from the position on the image.

(Ninth Embodiment) FIG. 22 is a diagram showing a configuration of a position detecting device according to a ninth embodiment of the present invention for detecting the positions of one or more objects in a wide range. A large number of omnidirectional imaging devices 1 of any of the above-described first to fourth examples are arranged in a matrix within a range of position detection. With such a configuration, it is possible to accurately detect the position of one or a plurality of objects even in a very wide range.

(Tenth Embodiment) In the present invention in which the position of an object is detected by using a plurality of omnidirectional imaging devices, the detection range can be easily expanded by increasing the number of the omnidirectional imaging devices. It is preferable to provide a large number of omnidirectional imaging devices in order to reduce the detection error. When the number of omnidirectional imaging devices installed is increased in this manner, the number of omnidirectional images acquired also increases, and a larger number of omnidirectional images are acquired by the position detector 53 (main control unit 53a),
Computer 56, position detector 59 (main controller 59a)
And so on. As a result, real-time processing may not be performed.

The tenth embodiment has been made in order to eliminate such a possibility, and the object position detection processing is not performed intensively by one device, but is performed in a distributed manner. Be able to do it in parallel. FIG. 23 is a diagram showing the configuration of the object position detection system according to the tenth embodiment of the present invention. In FIG. 23, the four omnidirectional imaging devices 1 of any of the first to fourth examples described above are arranged in a predetermined detection space.
Is provided. Each of these four omnidirectional imaging devices 1 is provided with a position dispersion detector 71. These four position dispersion detectors 71 are connected to a LAN 72. The LAN 72 has one position center detector 7.
3 are connected.

Each position variance detector 71 includes a main controller 71
a, a ROM 71b, a RAM 71c, a net interface 71d, and the like. The main control unit 71a is specifically configured by a CPU, controls each of the above-described hardware units in the position dispersion detector 71, and
According to the computer program stored in the OM 71b, various software functions in estimating the position of the object are executed. The RAM 71c stores temporary data generated when executing software. The net interface 71 d controls the input and output of data between the other position dispersion detector 71 and the position center detector 73 via the LAN 72.

Next, the operation will be described. In the above-described first to ninth embodiments, all processes (S2 to S6) other than the acquisition of the omnidirectional image (S1) are performed by the central position detector 53, the computer 56, the position detector 59, and the like. However, in the tenth embodiment, the object position detection processing (S2 to S6) is performed separately for the omnidirectional imaging device 1 and the central side. That is, part of the object position detection processing (from S2 to S44 to S44) is performed by each position dispersion detector 71, and the remaining processing (from S4 to S45) is performed by the position center detector 73. At this time, the detection data (azimuth) of the other omnidirectional camera 1 required for the processing is L
Sent via AN72.

In the tenth embodiment, since the object position detection processing is performed in a distributed manner, the processing can be performed in parallel. Also, the detection processing can be performed in real time.

Although the description has been given of several types of systems to which the position detecting method and the position detecting device of the present invention are applied, these are merely examples, and a plurality of omnidirectional imaging devices for acquiring images in all directions in the circumferential direction are used. Of course, other systems that utilize the detection of the position of an object are also possible.

In the example described above, the case where the position of an object is detected using a plurality of omnidirectional imaging devices has been described. However, by using the system of the present invention, one or more objects at an arbitrary position can be detected. Since an image can be obtained, it is of course possible to detect visual features such as the shape and color of the object in addition to the positional information. By incorporating this visual feature, it is possible to further improve the position detection accuracy of the object.

(Eleventh Embodiment) In the above-described example, a case has been described in which an omnidirectional photographing device for acquiring an omnidirectional image is used. The present invention can be implemented in the same manner by using. FIG. 24 is a diagram showing a configuration of a position detecting device for detecting the positions of one or a plurality of objects according to an eleventh embodiment of the present invention. In FIG. 24, a region surrounded by a broken line is a rectangular detection region. Near the four corners of the detection area, one wide-angle camera 101 is provided.

The wide-angle photographing device 101 includes a transparent cylinder 110 made of glass or plastic material, and a cylinder 110.
Wide-angle lens 111 and CCD element 1 housed inside
And 12. The outer peripheral surface of the cylindrical body 110 in front of the wide-angle lens 111 has a light shielding material 113 except for a part of the light incident portion.
Covered in. Instead of providing the light-blocking material 113 on the cylindrical body 110 as in this example, the light-blocking material may be provided directly on the surface of the wide-angle lens 111 to obtain a wide-angle image in a plane.

With such a configuration, each wide-angle camera 10
1 focuses light from a predetermined height at a light incident portion of a cylindrical body 110 on a CCD element 112 via a wide-angle lens 111, for example, in a circumferential direction including two objects A and B existing in a detection area. Is obtained in a wide-angle range (a range slightly larger than 90 °). This wide-angle range to be photographed completely includes a rectangular detection area.

Each wide-angle photographing device 101 is connected to a position detector 53 similar to that of the first embodiment.
The wide-angle image at the predetermined height position acquired at 01 is input to the position detector 53. In the position detector 53, the same processing as the processing on the image acquired by the omnidirectional imaging device 1 is performed on the image acquired by the wide-angle imaging device 101, and the objects A and B are processed.
Is detected.

[0099]

As described above, in the position detecting method according to the first aspect of the present invention, each of the plurality of image acquiring means acquires a circumferential image including the object to be detected, and the position of the object is determined based on the acquired images. Since the image is detected in all directions, a mechanism for driving the image acquisition means is not required because all directions are photographed at a time. Even with a small number of image acquisition means, the feature (color, shape, etc.) of the object is used. Instead, the position of the object can be accurately detected.

According to the position detecting method of the second invention, a difference image between the reference image not including the object and the image including the object is obtained, and the position of the object is detected from the difference image.
An object existing in an arbitrary direction can be easily identified by the background difference, and the azimuth of the object with respect to each image acquisition means can be easily obtained.

In the position detecting method according to the third aspect of the present invention, a moving object is tracked by acquiring an image in the circumferential direction including the object with time, so that a tracking result of the object having no time lag can be obtained. it can.

In the position detecting method of the fourth invention, the position of the object is detected in consideration of the error of the azimuth angle to the object and the error of the apparent size of the object in the image. The position of an object such as a human having a substantially circular cross section can be accurately detected.

In the position detecting device according to the fifth aspect of the present invention, a plurality of image obtaining means obtain circumferential images including the object to be detected, and the detecting means analyzes the obtained images to detect the position of the object. Therefore, a mechanism for driving the image acquisition means is not necessary because all directions are photographed at once, and only a small number of image acquisition means can be used without using the features (color, shape, etc.) of the object. In addition, the position of the object can be accurately detected.

In the position detecting device according to the sixth aspect of the invention, the necessary light not regulated by the regulating member is reflected by the rotationally symmetric light reflector, and then incident on the image pickup device to obtain an image including the object. Since unnecessary light is regulated by the regulating member, an image on an unnecessary plane cannot be obtained, and an image including an object can be reliably obtained.

In the visual characteristic detecting method according to the seventh aspect of the present invention, a plurality of image obtaining means obtain circumferential images including the object to be detected, and analyze the obtained images to obtain visual characteristics of the object. (Shape, color, etc.) is detected, so that all directions, that is, a 360-degree range or a wide-angle range, are photographed at a time. Therefore, a mechanism for driving the image acquisition means is unnecessary. Even so, the visual characteristics of the object can be accurately detected.

In the pointing system of the eighth invention,
Each of the plurality of image acquisition units acquires a circumferential image including the indicator contacted with the contacted object, and the detection unit detects the position indicated by the indicator based on the acquired images. Therefore, position detection can be performed with a simple configuration without providing a scanning mechanism and a driving unit.

In the monitoring system according to the ninth aspect, an image in the circumferential direction in a predetermined plane area is obtained by each of the plurality of image obtaining means, and the detection means detects intrusion of an object based on the obtained plurality of images. With this configuration, the intrusion of an object can be detected with a simple configuration without providing a scanning mechanism and a driving unit.

[Brief description of the drawings]

FIG. 1 is a diagram showing a configuration of a first example of an omnidirectional imaging device used in the present invention.

FIG. 2 is a diagram showing a configuration of a second example of an omnidirectional photographing device used in the present invention.

FIG. 3 is a diagram showing a configuration of a third example of an omnidirectional imaging device used in the present invention.

FIG. 4 is a diagram showing a configuration of a fourth example of an omnidirectional photographing device used in the present invention.

FIG. 5 is a flowchart showing a procedure of an object position detection method of the present invention.

FIG. 6 is a diagram showing a correspondence between stereo vision and an object.

FIG. 7 is a flowchart showing a procedure for estimating the position of an object (subroutine of step S4 in FIG. 6).

FIG. 8 is an explanatory diagram of a method of estimating a position of an object in consideration of an observation error.

FIG. 9 is an explanatory diagram of a method of estimating the position of an object by binocular stereo vision in consideration of an observation error.

FIG. 10 is a diagram showing a state of erroneous detection of an object by three-lens stereo vision.

FIG. 11 is an explanatory diagram of a method of estimating the position of an object by simplified N-eye stereo vision.

FIG. 12 is an explanatory diagram of an error correction method in a simplified position estimation method of an object by N-eye stereo vision.

FIG. 13 is an explanatory diagram of an error correction method in a simplified method of estimating an object position by N-eye stereo vision.

FIG. 14 is a diagram showing a configuration of a pointing system according to the first embodiment.

FIG. 15 is a diagram illustrating a configuration of a pointing system according to a second embodiment.

FIG. 16 is a diagram illustrating a configuration of a monitoring system according to a third embodiment.

FIG. 17 is a diagram illustrating a configuration of a maintenance system according to a fourth embodiment.

FIG. 18 is a diagram showing a configuration of a shape measuring system according to a fifth embodiment.

FIG. 19 is a diagram showing a configuration of a performance system according to a sixth embodiment.

FIG. 20 is a diagram illustrating a configuration of an object detection system according to a seventh embodiment.

FIG. 21 is a diagram showing a configuration of a human tracking system according to an eighth embodiment.

FIG. 22 is a diagram illustrating a configuration of a position detection device according to a ninth embodiment.

FIG. 23 is a diagram illustrating a configuration of an object position detection system according to a tenth embodiment.

FIG. 24 is a diagram illustrating a configuration of a position detection device according to an eleventh embodiment.

[Explanation of symbols]

 1, 1a, 1b, 1c Omnidirectional photographer 10a Light incident part 11 Convex mirror 13 Light shielding material 16, 33 Camera 19, 35 Lens 20, 36, 112 CCD element 21 Concave mirror 31 Spherical mirror 32a Guide part 41 Linear body 51 Panel 52 Instruction Pens 53, 59 Position detectors 53a, 59a, 71a Main controllers 53e, 59c, 71b ROM 54 Screen 55 Projector 56 Computer 57 Wall 58 Painting 59f Alarm 101 Wide-angle camera 111 Wide-angle lens

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Takeshi Maeda 4-27-2 Kodai, Seika-cho, Soraku-gun, Kyoto Etoile B101 (72) Inventor Shinji Oki 1-125-609 Minamishinmachi 1-chome, Matsubara-shi, Osaka (72) Inventor Nobuo Yamato 2-2-18 Shimodera, Naniwa-ku, Osaka-shi, Osaka F-term (reference) 2F065 AA03 BB01 DD02 DD03 FF04 JJ26 LL19 QQ24 QQ31 5B068 AA04 AA22 AA32 BB18 BC05 BD09 BD17 BD25 BE08 CC17 CD06 DE12

Claims (9)

[Claims] 1. A method for detecting a position of one or a plurality of objects in a plane, wherein a plurality of image acquiring means provided with a light incident portion at a position corresponding to the plane includes a peripheral direction including the object. A position detection method comprising: obtaining an image of the object; and detecting a position of the object based on the obtained images. 2. The image acquisition device according to claim 1, wherein the position of each of the image acquisition devices is fixed, and the first image in the circumferential direction not including the object acquired by each of the image acquisition devices and the image acquisition device each include: The position detection method according to claim 1, wherein a difference image from a second circumferential image including the object obtained by the calculation is obtained, and a position of the object is detected according to the obtained difference image. 3. The position detecting method according to claim 1, wherein each of the image obtaining means obtains an image in a circumferential direction over time, and detects a position of the object over time. 4. When the cross-sectional shape of the object by the plane is substantially circular, an error in the azimuth angle with respect to the object in each of the image acquisition units, and the object acquired by each of the image acquisition units The position detection method according to claim 1, wherein the position of the object is detected in consideration of an error in an apparent size of the image. 5. An apparatus for detecting a position of one or a plurality of objects in a plane, wherein the apparatus is provided at a position corresponding to the plane and acquires a circumferential image including the object. And a detecting means for detecting a position of the object based on a plurality of images acquired by each of the image acquiring means. 6. Each of the image acquisition means includes a light reflector having a rotationally symmetric shape, an image pickup device provided at a position facing the light reflector, and light incident on the light reflector. The position detecting device according to claim 5, further comprising a regulating member that regulates a range. 7. A method for detecting a visual characteristic of one or more objects in a plane, wherein the object is included by each of a plurality of image acquisition means provided with a light incident portion at a position corresponding to the plane. A visual feature detection method comprising: acquiring a circumferential image; and detecting a visual feature of the object based on the acquired plurality of images. 8. A pointing system for detecting a position pointed by an indicator, wherein a contacted object contacted by the indicator and a circumferential image including the indicator contacted by the contacted object are displayed. A pointing system, comprising: a plurality of image acquiring means for acquiring; and a detecting means for detecting a position indicated by the pointer based on a plurality of images acquired by each of the image acquiring means. 9. A system for monitoring the intrusion of an object into a predetermined plane area, a plurality of image acquisition means for acquiring images in the circumferential direction in the plane area, and a plurality of image acquisition means acquired by each of the image acquisition means. A monitoring system, comprising: detection means for detecting intrusion of the object based on an image.