JPH0614356B2 - Information extraction method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Outline] On a mapping surface, a point corresponding to each photographing position is obtained along a moving direction of a camera with a mapping point indicating a pole of a moving locus of a photographed image as a reference point. The spherical mapping function (great circle) corresponding to the mapping point and the point where the captured image intersects at each shooting position are obtained, and a spherical mapping function including both corresponding points at each shooting position is generated.
An information extraction method for extracting three-dimensional information of a photographed image based on the obtained intersections.

[Industrial application field]

The present invention relates to an information extraction method for extracting three-dimensional information such as line segments and points from an original image by spherical mapping.

In recent years, object recognition technology has been actively developed in order to realize a function similar to that of the human eye by a machine. In the recognition of such an object, a method of recognizing the object by capturing the object as an image and extracting information such as the feature of the object from the image is used. Therefore, it is generally necessary to capture an original image by capturing an image of an object with an image capturing unit such as a television camera and perform electrical processing based on this image to extract information.

On the other hand, such an original image often contains noise or has bleeding or blurring, and it is necessary to remove these noise components and extract the structure lines or the like constituting the original image.

[Conventional technology]

As a method of extracting such an image, an information extraction method by mapping has been conventionally known. A method using a spherical map as an information extraction method using this map (JP-A-60-218183).
Issue gazette).

With this spherical mapping, the original image is projected onto a spherical surface (spherical projection), and as shown in FIG. 5 (a), that point is projected for each point P of the original image projected onto the spherical surface. A great circle C, which is the center of the circle, is drawn, this is expanded on the mapping surface as shown in FIG. 12 (b), and information is extracted from the mutual relation of the drawn great circle group.

By following such a procedure, a normal vector of a plane including each point of the original image and the center of the spherical surface can be obtained, a line segment can be extracted at the same time, and a line segment that is discontinuous or distorted can be reproduced. Furthermore, by mapping the change information of the projection on the spherical surface to one point on the spherical surface, it is possible to extract the movement locus of the point-like object that moves linearly.

[Problems to be solved by the invention]

However, in the conventional information extraction method using spherical mapping, only the existence of line segments or the direction of line segments can be extracted, and many exist in the measurement target space required for actual work using a robot or the like. It was not possible to extract 3D information such as possible straight lines, curves, planes, and curved surfaces.

In view of the above-mentioned conventional problems, an object of the present invention is to provide an information extraction method capable of extracting a distance to a captured image such as a point, a line segment, or a surface by a spherical mapping.

[Means for solving problems]

An object of the present invention is to provide a camera for photographing a space to be measured, a means for moving the camera in a three-dimensional direction, a means for generating a spherical mapping function for a photographed image obtained by the camera, And a mapping surface to which a great circle is added by a spherical mapping function, images are taken at different positions when the camera is moved linearly, and a spherical mapping function is generated for the taken images at the different positions. Writing on the surface, extracting the intersection Σ at which a plurality of great circles intersected by the spherical mapping function on the mapping surface obtained by the writing, and extracting the pole of the locus of movement of the photographed image, and extracting the pole on the mapping surface. As a reference point, a first point group T ⁰ , T ^1, ... T corresponding to the respective photographing positions along the moving direction of the camera
ⁱ ) and a second point group S _h ⁰ , S _h at which the great circle by the spherical mapping function corresponding to the pole intersects the photographed image obtained at each photographing position
_h ¹ ... S _h ⁱ is obtained, a spherical mapping function including a set of points corresponding to each photographing position in the first and second point groups is generated, written on the mapping surface, and obtained by the writing. Achieved by an information extraction method characterized by extracting an intersection Q of a great circle by a spherical mapping function on the mapping surface corresponding to each captured position, and extracting three-dimensional information of the captured image from the position of the intersection. To be done.

1 to 3 are diagrams for explaining the basic principle of the information extraction method according to the present invention.

First, consider a method of measuring the distance from the camera T to the point P on the two-dimensional plane.

FIG. 2 is a diagram showing the movement of the camera with respect to the point P.

Now, assuming that the camera is at the position T ⁰ , the point P is seen from the camera in the direction of the line of sight S _h ⁰ . Next, when the camera is moved to T ¹ , the direction in which the point P is viewed becomes S _h ¹ . When the camera is linearly moved to T ⁰ , T ¹ , T ² , ...
The direction in which the point P is visible changes to S _h ⁰ , S _h ¹ , S _h ² , .... When a straight line is drawn from the position of the camera at each point in the direction in which the point P can be seen, these straight lines intersect at one point P.

Therefore, if the intersection point P is obtained, the distance from the initial position T ⁰ of the camera to the target point P is given by the length of the line segment T ⁰ P.

A case where the same operation is performed on the spherical surface will be described below.

In FIG. 3, the right plane is the same as in FIG.
It is placed perpendicular to the line segment OΣ. Also, the correspondence from the flat surface to the spherical surface is shown by a dotted line.

The moving direction of the camera is vector V. The camera is Δx ₀
It is assumed that the point P moves and looks like P ⁰ , P ¹ , P ² , ... With each advance, and the projections on the spherical surface are S _h ⁰ , S _h ¹ , S _h ² ,.
There is a point Σ in the direction perpendicular to the end point v of the vector V. This point Σ
Is obtained as the intersection of the spherical mapping functions for S _h ⁰ , S _h ¹ ... S _h ⁱ . The time axis (τ axis) is taken on the 1/4 circle from the point Σ to the point v of the great circle passing through the points v and Σ, and the point Σ
Be τ = 0, that is, the point T ⁰ .

Hereinafter, the point where the length of the arc from the point P ⁰ is τ = tan ⁻¹ (iη), i = 1, 2, ... (1) is defined as T ⁱ . Here, η is a scale factor represented by η = Δx ₀ / R ₀ (2). R ₀ is an arbitrary value called a distance coefficient. This procedure corresponds to multiplying the interval of T ⁰ , T ¹ , T ² , ... On the right plane of FIG. 3 by 1 / R ₀ and plotting it on the spherical surface. As can be seen by setting i → ∞ in the above equation (1), the point v is the point at infinity on the τ axis.

Next, when the points T _i and S _h ⁱ are connected by a great circle for i = 0, 1, 2, ..., The great circles intersect at a point Q. From this, the distance from the initial position T ⁰ of the camera to the point P is R
It is given by ₀ tan (length of arc T ⁰ Q).

On the other hand, the distance to the line segment can be obtained in the same manner as the method for obtaining the distance to the above point, which will be described below.

The line segment to be measured is L. Pay attention to the foot of the perpendicular line drawn from the camera to the line segment L. As shown in FIG. ¹ , it is assumed that the line segment L moves and appears as L ⁰ , L ¹ , L ² , ... Each time the camera advances by Δx _0. At each time point, the line segment L is projected on a spherical surface. The points S ⁰ , S ¹ , S ² , ... Which are the poles of the circle and the point SS which represents the three-dimensional direction of the line segment L are obtained using the technique disclosed in Japanese Patent Laid-Open No. 60-218183. The point SS is obtained as the intersection of the spherical mapping functions with respect to the points S ⁰ , S ¹ , S ² ... On the great circle having the point SS as a pole, the points S ⁰ , S ¹ , S ² , ... Are perpendicular to the points S _h ⁰ , S _h ¹ , S _h ² ,.
Take This sequence of points is the projection of the perpendicular line drawn from the camera to the line segment L at each time point on the spherical surface of the feet P ⁰ , P ¹ , P ² ,.

The intersection point of the great circle passing through the end point v and the point SS of the camera movement direction vector V and the great circle in which the point sequences S _h ⁰ , S _h ¹ , S _h ² , ...
And The τ axis is set on the 1/4 circle from the point SS to the point v ′, and the point SS is τ = 0, that is, T ⁰ . Below, point T ⁰
Let T ⁱ be the point where the length of the arc from is τ = tan ⁻¹ (i η sin γ), i = 1, 2, ... (3).

η is the scale factor of the above equation (2), and γ is the angle formed by the point v and the point SS. When a point T ⁱ and a point S _h ⁱ are connected by a great circle for i = 0, 1, 2, ..., These great circles intersect at a point Q. From this, the distance from the initial position of the camera to the line segment L is given by R ₀ tan (the length of the arc T ⁰ Q).

〔Example〕

An embodiment of an information extraction method according to the present invention will be described in detail below with reference to the drawings.

(1) Description of Configuration of Embodiment FIG. 4 is an overall configuration diagram of a spherical mapping device to which the present invention is applied.

In the figure, 30 is a microprocessor (hereinafter referred to as MPU), 31 is a spherical camera, 32 is an original image memory, 34
Is a moving direction measurement data creation unit, 35 is a moving direction determination unit, 36, 37, 38 are mapping processors, 39, 40,
Reference numeral 41 is a mapping memory, and 42 is a parallel interface.

In the above configuration, the MPU 30 controls the spherical camera 31, the original image memory 32, the moving direction measurement data creation unit 34, the moving direction determination unit 35, and the like. The spherical camera 31 photographs an object and obtains image data. In the polar coordinate (r, θ) format, and the original image memory 32 stores the image data in the polar coordinate format output from the spherical camera 31 and subjected to the contour extraction processing in association with each other in a grid pattern. There is a storage area that is divided into N in the longitude direction and M in the latitude direction, and the moving direction measurement data creation unit 34 sets any of a plurality of points on the measurement target space to the attention points P ₀ ^t , P ₁ ^t , P ₂
It is set as ^t ..., for example, when the vertexes at which a plurality of line segments intersect in the original image from which contours have been extracted are set as points of interest, or when a plurality of light spots are irradiated on the measurement target space. Of the original image and the original image when not irradiated, and sets the coordinates of the obtained light spots as the points of interest. The movement direction determination unit 35 determines which coordinate the intersection point developed in the mapping memory 39 corresponds to. The moving direction of the spherical camera 31 is determined based on whether the spherical camera 31 is stored. The mapping processor 36 generates great circle information for the image data of each pixel stored in the original image memory 32. A mapping memory for sequentially storing mapping results, a mapping memory 39 for transferring and storing peak points (intersection points) which are mapping results in the mapping processor 36, and a mapping processor 37 for a mapping memory. 9 for generating great circle information for the pixel having the peak point stored in 9 and sequentially storing the mapping result in a mapping memory provided therein. The mapping memory 40 is a peak point which is the mapping result of the mapping processor 37. Is transferred and stored from the internal memory in the mapping processor 37, and the mapping processor 38 generates great circle information for the pixel having the peak point stored in the mapping memory 40 and the mapping provided therein. A memory for sequentially storing the mapping result, a mapping memory 41 for transferring and storing the peak points stored in the mapping memory in the mapping processor 38, and a parallel interface 42 for the MPU 30 or the like in the mapping memories 39, 40, 41. To access.

(2) Description of Operation of Embodiment With the above-described configuration, a method of extracting three-dimensional information of a captured image of line segments, points, etc. will be described.

(a) Measurement of direction of line segment First, the spherical camera 31 captures an image of a measurement target space under the control of the MPU 30 and outputs image data of which contours have been extracted in polar coordinates (r, θ) format. Then, the MPU 30 writes the image data (coordinate position and density value) from the spherical camera 31 in the original image memory 32 in pixel units.

Next, the mapping processor 36 generates a spherical mapping function (great circle) for the image data of each image in the original image memory 32, and stores the peak extraction result in the mapping memory 39.

Therefore, the mapping memory 39 stores the intersections of the great circles generated corresponding to the points constituting each line segment obtained by contour extraction, that is, the poles (S points) of the line segment.

After the spherical mapping function generation process in the mapping processor 36 is completed, the MPU 30 operates a three-dimensional positioning device (not shown) equipped with the spherical camera 31 to move the spherical camera 31 by a predetermined pitch. Then, the processings 1 and 2 are executed, and thereafter, the spherical camera 31 is moved straight (translationally) to photograph a predetermined number of frames.

Thereafter, the mapping processor 37 causes the mapping memory 39
A spherical mapping function is generated for each pixel of each peak point stored in, and the peak extraction result is stored in the mapping memory 40. Therefore, in the mapping memory 40, information indicating the direction of the line segment existing in the measurement target space is the intersection point (SS point).
Is stored as

(b) Measurement of Camera Moving Direction Here, a method of measuring the moving direction of the camera, which is necessary when performing the three-dimensional information extraction processing of the line segment described later, will be described.

First, the camera moving direction measuring method will be described below.

FIG. 5 is an explanatory diagram of the basic principle of the camera moving direction measuring method. In the figure, P ₀ ^t , P ₁ ^t and P ₂ ^t (t is time series information, t = 0, 1, 2, ...) Is an arbitrary point of interest in the space to be measured, and the spherical surface A is a spherical surface memory for recording information of a camera which images the points of interest P ₀ ^t , P ₁ ^t , and P ₂ ^t .

Now, when the plurality of points of interest P ₀ ^t , P ₁ ^t , and P ₂ ^t which are stationary while the camera is linearly moved, it seems that the plurality of points are linearly moved in parallel.

Therefore, for example, as shown in FIG. 5, a plurality of points are spherically mapped (great circle generation) at each time point, and the results are accumulated in the spherical memory A. Column Σ ₀ ,
Σ ₁ , Σ ₂ , ... Are obtained. When a spherical map is applied to this Σ point sequence and the great circle information is written in the spherical memory A, the intersection v of them is obtained.

Here, the vector V connecting the intersection v and the center of the sphere is parallel to a plurality of straight lines connecting the point sequences projected on the sphere due to the nature of the sphere.

On the other hand, since the plurality of straight lines projected on the spherical memory A and the moving direction of the camera coincide with each other, the moving direction of the camera coincides with this vector V.

Therefore, the three-dimensional direction of the camera can be known from the address of the intersection v on the spherical memory A.

This operation will be described below.

First, the same processing as the processing of (a) described above is executed on the image data that is not subjected to the contour extraction processing.
The image data from the spherical camera 31 is the original image memory 3
It is not stored in 2, but is stored in the first internal memory in the moving direction measurement data creation unit 34.

Then, before performing the process of (a) described above,
The MPU 30 keeps the spherical camera 31 stationary,
A spot light irradiation unit (not shown) is operated to irradiate a plurality of spot lights on the same measurement target space.

Then, the target space is photographed by the spherical camera 31 in the same manner as the processing of the item (a). Then, the obtained image data that has not been subjected to the contour extraction processing is stored pixel by pixel in the corresponding storage area of the second internal memory in the moving direction measurement data creation unit 34.

In the moving direction measurement data creation unit 34, the first and second
When the image data not irradiated with the spot light and the image data irradiated with the spot light are stored in the internal memory of, the difference between the two is calculated.

As a result, only the point irradiated with the spot light is taken out as a difference, and therefore the moving direction measurement data creation unit 34 extracts the coordinate value (r, θ) of this point and further stores it in the third internal memory. Write the coordinate value of this point.

Thereafter, in the above-mentioned item (a), the MPU 30 gives an instruction to the three-dimensional positioning device (not shown) on which the spherical camera 31 is mounted so as to linearly move by a predetermined pitch,
By sequentially repeated to perform the processing of ~ each changes the position of the three-dimensional space of the spherical camera 31 is information for measuring a moving direction of the camera, the point of interest P ₀
Time-series information ^{_{P 0 1, P 0 2,}} P 0 3 against, ..., time-series information P _{¹ 1} point of attention ^{_{P 1, P 1 2, P}} 1 3, ..., time series for the target point P ₂ information P ₂ ¹ , P ₂ ² , P ₂ ³ , ... Are written as shown in FIG.

In this way, measurement data for measuring the three-dimensional movement direction of the spherical camera 31 is obtained.

Next, the moving locus of each target point is obtained from the measurement data thus obtained.

First, the MPU 30 writes the measurement data in the original image memory 32 from the internal memory in the moving direction measurement data creation unit 34.

Next, the mapping processor 36 is operated to generate a spherical mapping function (great circle) for the pixel of each point of interest written as shown in FIG. 6 and write it in the internal mapping memory.

FIG. 6 is a diagram showing a state in which great circles corresponding to respective points of interest are written in the internal mapping memory. For the point of interest P ₀ ¹ , the great circle L _P0 ¹ and for the point of interest P ₁ ¹ are shown. It is written as a great circle L _P1 ¹ , ...

Then, the mapping processor 36 causes the intersection points Σ ₀ , Σ ₁ , Σ of the great circles L _P0 ¹ ... Written in the internal memory in this way.
₂ are extracted, and the intersections Σ ₀ , Σ ₁ , Σ ₂ , ...
Write in 9.

As a result, the intersection, which is the pole of the locus of the point of interest that moves in a straight line, is obtained.

Further, the moving direction of the camera is measured from the information indicating the moving locus thus obtained.

First, as shown in FIG. 7, the intersections Σ ₀ , Σ ₁ ,
A spherical mapping function (great circle) is generated by the mapping processor 37 for the pixels of Σ ₂ , ... And written in the internal mapping memory.

FIG. 7 is a diagram showing a state in which the great circles for the intersections Σ ₀ , Σ ₁ , Σ ₂ , ... Are written in the internal memory. For the intersections Σ ₀ , the great circle L ₀ , the intersections Σ ₁ , Is written as a great circle L ₁ and an intersection Σ ₂ is written as a great circle L ₂ .

Then, the mapping processor 37 writes in the mapping memory 40 the intersection v of the great circles L ₀ , L ₁ , ... Written in the internal memory in this way.

The MPU 30 operates the moving direction determination unit 35 when the intersection v is written in the mapping memory 40.

The moving direction determination unit 35 has a parallel interface 42.
The intersection point v in the mapping memory 40 is read via the, and the moving direction of the spherical camera 31 corresponding to the read position is determined.

That is, the mapping memory 40 stores the image data in the polar coordinate (r, θ) format, and the coordinate position itself of the image data represents the position of the point where the center position of the sphere intersects with the spherical surface. can do.

Then, the MPU 30 holds this moving direction data in the internal memory.

In the process of measuring the moving direction of the camera, the contents of the mapping memory 39 and the contents of the mapping memory 40 storing the information indicating the direction of the line segment need to be transferred to another memory (not shown). There is. Or spherical camera 3
If the distance measurement processing is performed after the movement direction measurement processing 1 is completed, this is not necessary.

(c) Line segment distance measurement First, a plurality of line segment directions exist in the measurement target space, and a plurality of intersections (SS) corresponding to each line segment direction are stored in the mapping memory 40.

Therefore, when measuring the distance of a line segment, since the direction of one line segment among the plurality of line segments is targeted, the MPU 30 searches the contents in the mapping memory 40 via the parallel interface 42 and Any of the intersections (see Fig. 8
Select SS point). Then, it instructs the map processor 38 about the coordinates of this point.

The mapping processor 38 generates a corresponding spherical mapping function (great circle) based on this coordinate value and stores it in the internal memory. Then, the MPU 30 intersects with the great circle information in the internal memory of the mapping processor 38 and the pole (S) of the line segment stored in the mapping memory 39 (S ⁰ , S ¹ , S ² , ... In FIG. 8). ) Is extracted.

As a result, line segment information for each frame for the line segment 1 having the line segment direction to be measured (l ⁰ , l ¹ , l ² ,
..) are extracted (S ⁰ , S ¹ , S ² , ... In FIG. 8).

On the other hand, as shown in FIG. 9, consider a spherical surface having the obtained intersection point SS as the north pole. A sphere with the intersection point SS as the north pole can be easily obtained by coordinate transformation. On this spherical surface, a point C ^t representing the position of the camera is obtained for each frame t.
This camera position is obtained by the above equation (3).

That is, the MPU 30 is configured with the preset spherical camera 31.
R ₀ and the moving direction v of the camera that determines the reduction ratio when expressing the moving pitch Δ _{x0 of}
C ^t (t = 1, 2, ...) For each frame t is obtained using the angle γ formed by the point SS obtained from the above and stored in the internal memory.

The coordinates T ^t (▲ T ^t _α ▼, T ^t (▲ T ^t _α ▼) of the point C ^{t obtained} by projecting the position of the spherical camera 31 on the mapping memory 40 at each time on the spherical surface.
▲ T ^t _β ▼) is the moving direction v of the spherical image camera on the original image memory 32 and the coordinates of the intersection SS are v (v _α ,
v _β ), SS (SS _α , SS _β ), However, However, η = Δx ₀ / R ₀ γ = cos ⁻¹ {cosv _β cosSS _β + sinv _β sinSS _β cos (v _α −SS _α )} Δ _x0 ; camera moving step R ₀ ; distance coefficient.

Next, as shown in FIG. 9, a point S _h ^t representing the direction in which the foot of the perpendicular to each line segment l ⁰ , l ¹ , S ² , ... Can be seen on the sphere having the intersection SS as the north pole for each t. Seeking, MPU30
Stored in the internal memory of.

That is, when the intersection point SS is the north pole, the point S _h ^t is equal to the coordinates when the poles S ⁰ , S ¹ , S ² , ... Of the line segment are rotated by 90 degrees about the intersection point SS. Can be obtained by the angle of the pole of each line segment from +90 degrees.

That is, the coordinates ▲ S ^t _h ▼ (▲ S ^t _hα ▼, ▲ S ^t _hβ ) ▼ of the point that _projects the foot of the perpendicular to the line segment on the mapping memory 40 is However, (▲ S ^t _α ▼, ▲ S ^t _β ▼); S at time t of each line segment
Coordinates of points on cost memory 32 Required by.

Next, the MPU 30 poles b of the great circle (l _s0 , l _s1 , l _s2 ... In FIG. 9) connecting C ^t and S ^t thus obtained.
^t is calculated for each frame t.

the coordinates of the ^{b t b t (▲ b t} α ▼, ▲ b t β ▼) if, ▲ b ^t _β ▼ ＝ ｜ cot ^-1 (tanT ▲ ^t _β ▼ ・ sin ｜ ▲ S ^t _hα ▼-
▲ T ^t _α ▼ |) |.

Then, the coordinates (b ⁰ , b ¹ , b ² , ... In FIG. 9) of the pole b ^t for each frame obtained in this way are given to the mapping processor 38.

The mapping processor 38 thereby causes each pole b ^t to be given.
A spherical mapping function (great circle of l _s0 , l _s1 , l _s2 , ... in Fig. 9) corresponding to is generated, written in the internal memory, and the coordinates (intersection Q) of the peak extraction result are written in the mapping memory 41. Put in.

The MPU 30 reads the coordinates (Q _α , Q _β ) of the peak point stored in the mapping memory 41 and obtains the line segment distance R by the following formula.

R = R ₀ tan Q _β However, R ₀ ; distance coefficient From this point onward, the MPU 30 sequentially repeats the above-described processes (1) to other intersections SS in the mapping memory 40 to thereby obtain a straight line in the measurement target space It can measure all the distances of line segments.

(d) Measurement of point distance In order to measure the distance of a point, first, the spherical camera 31 is operated under the control of the MPU 30 to photograph the space to be measured, and the contour is extracted in polar coordinates (r, θ) format. Output the image data. Then, the MPU 30 stores the image data (coordinate position and density value) from the spherical camera 31 in the original image memory 32 in pixel units.

Next, the MPU 30 operates the three-dimensional positioning device (not shown) on which the spherical camera 31 is mounted to operate the spherical camera 3
1 is moved straight (translated) by a predetermined pitch in a state where its posture does not change.

Then, the above-described processing is executed, and thereafter, the spherical camera 31 is linearly moved to photograph a predetermined number of frames.

After that, the mapping processor 36 causes the original image memory 32
A spherical mapping function (great circle information) is generated for each pixel of each frame stored in, and the peak extraction result is stored in the mapping memory 39.

Therefore, the intersection Σ which is the pole of the locus of the point which moves the direct image is obtained in the mapping memory 39.

Here, a large number of points exist in the measurement target space, and a plurality of linear movement loci of the points exist in the original image memory 32. Therefore, when measuring the point distance, one of the plurality of trajectories is targeted, and the point depicting the trajectory is measured. Therefore, the MPU 30 stores the mapping memory 39 in the mapping memory 39 via the parallel interface 42. Explore,
An arbitrary intersection (point Σ in FIG. 10) is selected from the plurality of intersections, and the coordinates of this point are instructed to the mapping processor 37.

The mapping processor 37 generates a corresponding spherical mapping function based on this coordinate value and stores it in the mapping memory provided inside.

Then, the MPU 30 extracts an intersection (point P ⁰ , P ¹ , P ² , ... In FIG. 10) between the great circle information in the mapping memory in the mapping processor 37 and the point data stored in the original image memory 32. To do.

On the other hand, as shown in FIG. 11, consider a spherical surface having the selected intersection Σ as the north pole. A sphere with the intersection Σ as the north pole can be easily obtained by coordinate transformation. On this sphere,
A point C ^t representing the position of the camera is obtained for each frame t. This camera position is obtained by the above equation (1). That is, the MPU 30 uses a preset moving pitch Δ _x0 of the spherical camera 31, a constant R ₀ that defines a reduction ratio when expressing the distance in the measurement target space on the spherical surface, and a moving direction V of the spherical camera 31. C ^t (t = 1, 2,
…) Is calculated by the following formula and stored in the internal memory.

However, However, V (v _α , v _β ); camera moving direction Σ (Σ _α , Σ _β ); coordinates of intersection Σ η = Δx ₀ / R ₀ Δ _x0 ; camera moving step R ₀ ; distance coefficient.

Next, as shown in FIG. 11, a point P ^t ′ representing the visible direction of each point is obtained for each t on the spherical surface having the intersection Σ as the north pole, and stored in the internal memory of the MPU 30. That is, when the intersection point Σ is the north pole, a point P ^t ′ (P ▲ ^{t ′} _α ▼, P ▲) representing the visible direction of each point on the spherical surface.
^{t ′} _α ▼) is ▲ P ^{t ′} _α ▼ = mod π (cos ⁻¹ (sin ▲ P ^t _β ▼ sin (▲ P ^t _α ▼
-Σ _α )) where (▲ P ^t _α ▼, P ▲ ^t _β ▼); coordinates on the original image memory 32 at time t of each line segment. Required by.

Next, the MPU 30 poles b of the great circle (l ₀ , l ₁ , l ₂ , ... In FIG. 11) connecting C ^t and S ^t thus obtained.
^t is calculated for each frame t.

the coordinates of the ^{b t b t (▲ b t} α ▼, ▲ b t β ▼) if, _{^{▲ b t β ▼ = | cot}} -1 (tan ▲ T t β ▼ · sin | ▲ P t 'α ▼ -
∑T ^{t ′} _α ▼ |) |

Then, the coordinates (b ⁰ , b ¹ , b ² , ... In FIG. 11) of the pole b ^t for each frame thus obtained are given to the mapping processor 38.

The mapping processor 38 thereby causes each pole b ^t to be given.
Spherical mapping function corresponding to (the great circle of l ₀ , l ₁ , l ₂ ... in Fig. 11)
Is generated and written in the internal mapping memory, and the coordinates (intersection Q) of the peak extraction result are written in the mapping memory 41.

The MPU 30 reads the coordinates (Q _α , Q _β ) of the peak point stored in the mapping memory 41 and obtains the point distance R by the following formula.

R = R ₀ tan Q _β However, R ₀ ; distance coefficient From this point onward, the MPU 30 sequentially repeats the above-mentioned processes (1) to the other intersections Σ in the mapping memory 39 to thereby obtain points in the measurement target space. Can measure all distances.

Therefore, for example, when obtaining the distance to a curved surface or a flat surface, it is possible to obtain the distance to the surface by irradiating this surface with spot light and performing processing such as obtaining the point coordinates of this spot light. it can.

〔The invention's effect〕

As described above, according to the present invention, it is possible to obtain the distance to a point, a line segment, or a surface by a spherical map.

[Brief description of drawings]

FIG. 1 is an explanatory view of the principle of the three-dimensional information extraction method of the line segment by the spherical mapping, FIG. 2 is an explanatory view of the principle of the three-dimensional information measuring method of the point by the spherical mapping, and FIG. FIG. 4, FIG. 4 is an overall configuration diagram of a spherical mapping device to which the present invention is applied, FIG. 5 is an explanatory diagram of a basic principle of a camera moving direction measuring method, and FIG. FIG. 7 is an explanatory view of a state in which a great circle is written to an intersection, FIG. 8 is a view showing the relationship between the intersection SS, the pole S of the line segment, and the projected line segment 1, FIG. Is an explanatory view of the process of obtaining the distance of the line segment, and FIG. 10 does not show the relationship between the intersection point Σ and the imaging point P.
FIG. 11 is an explanatory diagram of a process for obtaining the distance between points, and FIG. 12 is an explanatory diagram of the principle of spherical mapping. In the figure, 30 is an MPU, 31 is a spherical camera, 32 is an original image memory, 34 is a moving direction measurement data creating unit, 35 is a moving direction determining unit, 36, 37 and 38 are mapping memories, 39 and 4
0 and 41 are mapping memories.

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Yusuke Yasukawa 1015 Kamiodanaka, Nakahara-ku, Kawasaki City, Kanagawa Prefecture Fujitsu Limited Examiner Shigekazu Ishii

Claims (1)

[Claims] 1. A camera for photographing a space to be measured, a means for moving the camera in a three-dimensional direction, a means for generating a spherical mapping function for a photographed image obtained by the camera, and the spherical surface. A mapping surface to which a great circle based on a mapping function is added, images are taken at different positions when the camera is moved linearly, and a spherical mapping function is generated for the captured images at the different positions to create a mapping surface. , And the poles of the locus of movement of the photographed image are extracted by extracting the intersection (Σ) at which a plurality of great circles by the spherical mapping function on the mapping surface obtained by the writing are extracted. The reference point is a first point group (T ⁰ , T ¹ ...
T ⁱ ), and a second point group (S _h ⁰ , S 2) at which the great circle by the spherical mapping function corresponding to the pole intersects the photographed image obtained at each photographing position
_h ¹ ... S _h ⁱ ) is obtained, a spherical mapping function including a set of points corresponding to each photographing position in the first and second point groups is generated and written in the mapping surface, and by the writing, Information extraction characterized by extracting an intersection (Q) of a great circle by a spherical mapping function on the mapping surface corresponding to each obtained imaging position and extracting three-dimensional information of the captured image from the position of the intersection. Method.