JP2006113001A - Three-dimensional measuring method and device by photogrammetry - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To enable more accurate correspondence of a measurement point in photogrammetry. <P>SOLUTION: Using wide angle images of respective reference position and inspection position photographed while having a short focal length and at least one telephoto image of a part of the photographing region of the wide angle image that is photographed in a state where the focal length is longer than that of the wide angle images, a composite photograph image obtained by combining the telephoto image and the wide angle image photographed from the same photographing position so that a subject image of the telephoto image matches a subject image of the wide angle image and displaying the combined image is produced every photograph position. A reference point commonly taken in the photographing region of the telephoto image of the composite photograph image is specified, the photograph position relation is calculated based on coordinate values of the reference points in the plurality of composite photograph images, and three-dimensional position information in the subject space of the measurement point is measured from the coordinate values of an arbitrary measurement point in the photograph image based on the photograph positional relation. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to associating measurement points in photographic analysis using two or more stereo images, and more particularly to a method and apparatus for performing three-dimensional solid measurement by computer analysis using digital photographic images.

  Conventionally, a technique for analyzing stereo images, in particular, computer analysis of a subject captured in a digital image, has been used in the field of photogrammetry, and is particularly used for creating topographic maps. .

  In the computer analysis of the stereo image, it is necessary to derive the positional relationship between at least two cameras and to associate the measurement points imprinted on the photographic images taken by the two cameras. Specifically, the measurement point imprinted in one photographic image is specified in the other photographic image, and the two-dimensional coordinates on the plurality of photographic images and the three-dimensional coordinates of the camera and the orientation thereof are used. The 3D coordinates of the measurement point are derived.

  In photogrammetry that restores the 3D of an object from a photo, the same point (corresponding point) of the subject is specified among two or more photos with parallax. The problem is how to calculate three-dimensional coordinates with higher accuracy depending on how accurately the corresponding points are associated.

The work of associating a measurement point, which is an arbitrary point in the subject, with a coordinate position on the photograph in a plurality of photographs is difficult to accurately associate by visual manual operation. Therefore, Japanese Patent No. 3316682 (Patent Document 1) discloses a technique performed by correlation processing such as pattern matching using a computer.
Japanese Patent No. 3316682

  However, there are various subjects for photogrammetry, and it is difficult to photograph close to the subject when used for investigation of cliff landslides. Further, when it is necessary to take a photo over a wide area, the following methods can be considered, but each has the following problems.

  First, in order to shoot a subject far away from the shooting position in a wide area, a wide-angle image taken with a camera with a short focal length is used to shoot a plurality of wide-angle images with parallax. There are three-dimensional survey methods using In this case, since a wide-angle image taken with a camera with a short focal length is used, the number of pixels of the photograph corresponding to the unit area of the subject is reduced, so that the region of the subject corresponding to one pixel of the photograph image is reduced. As a result, the feature in the image is blurred. This phenomenon is particularly noticeable when a photographic image is enlarged and displayed. For this reason, the corresponding points specified using two photographic images with parallax are difficult to accurately identify the position of the subject, and it is difficult to accurately associate the photos.

  Second, using a telephoto image taken with a camera with a long focal length, take multiple photos from the same shooting position so that part of the shooting range overlaps, and connect them together to create a single photo There is a method of performing a three-dimensional survey using a large connected image and using the composite image. In this case, the number of pixels of the photograph taken per unit area of the subject increases, and the area of the subject projected per pixel is narrow, so the features appearing in the photograph become clear and the above problem Is resolved.

  However, when joining adjacent telephoto images in the process of creating a concatenated image, two photos are joined at the stage of joining the photographic images to join together multiple photos with different optical axes. There will be a slight error between the two. In particular, in the case of a photographic image, the subject image is greatly distorted due to lens aberration in the edge region of the image, and a plurality of photographs are taken so that the subject is common to both ends, and the plurality of photographs are connected. In the case of matching, the deviation tends to increase. When this is repeatedly connected until a connected image in which the entire subject has been photographed is created, the final connected image will be cumulatively displaced from the original image position, resulting in highly accurate connection. It was difficult to obtain an image. That is, since the connected image itself has an error from the true image of the subject, the obtained survey result includes an error even when the corresponding point is correctly specified.

  Therefore, a technical problem to be solved by the present invention is to provide a measurement point association method and apparatus capable of performing association with higher accuracy in association of measurement points in photogrammetry.

  In order to solve the above technical problem, the present invention provides the following three-dimensional measurement method.

According to the first aspect of the present invention, a plurality of photographic images are used by using a plurality of photographic images obtained by photographing a subject with a camera disposed at a reference position and a camera disposed at an inspection position different from the reference position. Is calculated based on the coordinate value in the photographic image of the reference point that is commonly captured in the photographic image, and the photographic position relationship between the reference position and the inspection position is calculated. A three-dimensional measurement method for measuring three-dimensional position information of the measurement point in the subject space from the coordinate value of the measurement point,
As the photographic images photographed from the reference position and the inspection position, respectively, a wide-angle image photographed with a short focal length, and a focal distance obtained by photographing a part of the photographing region from the same photographing position as the wide-angle image are obtained. Using at least one telephoto image taken in a long state, a composite photo image obtained by combining and displaying the telephoto image in the wide-angle image taken from the same photographing position is created for each of the photographing positions, A composite photo image,
In the composite photo image created for each shooting position, select a reference point that is commonly shown in the shooting area,
Provided is a three-dimensional measurement method for calculating a photographing position relationship between the reference position and the inspection position based on coordinate values in the plurality of composite photograph images of the reference point.

According to the second aspect of the present invention, the composite photograph image selects four or more reference points that are common to the wide-angle image and the telephoto image taken from the same photographing position,
A third aspect of the first aspect in which a projection conversion equation is calculated based on the coordinate value of the reference point in the wide-angle image and the telephoto image, and the telephoto image is synthesized and displayed in the wide-angle image by the projection conversion equation. A dimension measurement method is provided.

  According to the third aspect of the present invention, the composite image is obtained by measuring the X-axis dimension and the Y-axis dimension of a circumscribed frame circumscribing a conversion region obtained by projectively converting the telephoto image into the wide-angle image based on the conversion formula. And displaying the converted telephoto image in the wide-angle image based on the X-axis scale and the Y-axis scale, which are respectively derived as a ratio to the X-axis direction dimension and the Y-axis direction dimension of the telephoto image, and the projective transformation formula. A three-dimensional measurement method according to the second aspect created by the above is provided.

上記式において、ｘは望遠画像中の任意の点のＸ座標、ｙは望遠画像中の任意の点のＹ座標、ｘ_１６は変換望遠画像のＸ座標、ｙ_１６は変換望遠画像のＹ座標、ＳｘはＸ軸スケール、ＳｙはＹ軸スケール、Ａ，Ｂ，Ｃ，Ｄ，Ｅ，Ｆ，ａ，ｂ，ｃはそれぞれ射影変換式の定数を示す。 According to a fourth aspect of the present invention, there is provided the three-dimensional measurement method according to the third aspect, wherein the converted telephoto image is created by converting each point of the telephoto image according to the following equation.

In the above formula, x is the X coordinate of an arbitrary point in the telephoto image, y is the Y coordinate of an arbitrary point in the telephoto image, x ₁₆ is X coordinate of the transformation telephoto image, the Y coordinate of y ₁₆ converts telephoto image, Sx is an X-axis scale, Sy is a Y-axis scale, A, B, C, D, E, F, a, b, and c are constants of the projective transformation formula.

According to the fifth aspect of the present invention, a plurality of photographic images are used by using a plurality of photographic images obtained by photographing a subject with a camera disposed at a reference position and a camera disposed at an inspection position different from the reference position. Is calculated based on the coordinate value in the photographic image of the reference point that is commonly captured in the photographic image, and the photographic position relationship between the reference position and the inspection position is calculated. A three-dimensional measurement apparatus that measures three-dimensional position information of the measurement point in the subject space from the coordinate value of the measurement point,
As the photographic images photographed from the reference position and the inspection position, respectively, a wide-angle image photographed with a short focal length, and a focal distance obtained by photographing a part of the photographing region from the same photographing position as the wide-angle image are obtained. Image input means for inputting photographic image data with at least one telephoto image taken in a long state;
Coordinate input means for inputting a coordinate value on the photographic image for an arbitrary point of the photographic image input by the image input means;
Image composition means for creating, for each photographing position, a composite photograph image obtained by compositing the telephoto image in the wide-angle image photographed from the same photographing position among the photographic image data input by the image input means;
In the composite photographic image created for each shooting position, the coordinate value input by the coordinate input means for the same position of the subject that is common in the shooting area, as the coordinate value of the reference point, Based on the coordinate value, a positional relationship calculating means for calculating the shooting positional relationship of the photographic image;
With respect to the measurement position of the subject in the composite photograph image, the coordinate value input by the coordinate input means is used as the coordinate value of the measurement point, and based on the information on the photographing position relation calculated by the positional relation calculation means. And a measurement point calculation means for calculating a three-dimensional coordinate in the subject of the measurement point reflected in the composite photograph image.

  According to the first aspect of the present invention, a telephoto image and a wide-angle image are used to apparently combine the telephoto image and the wide-angle image so that the subject image in the telephoto image matches the subject image in the wide-angle image. That is, the composite image displays a telephoto image for the matching subject image when the composite photographic image is displayed. That is, since the composite photographic image uses fine information of the telephoto image for a part of the region, the composite image has fine image information for the region. Then, by selecting the reference point for the shooting area obtained by combining the telephoto images, the same point of the subject can be selected with high accuracy, and the position information is calculated and the measurement point is three-dimensionally in the subject space. Position information can be measured.

  According to the second aspect of the present invention, even when there is a shift between the center of the optical axis of the wide-angle image and the telephoto image at the time of creating the composite photo image, the telephoto image is subjected to projective transformation, so that the telephoto image is obtained. The images can be transformed as if they were taken from the optical axis of the wide-angle image, and the subject images of both photographic images can be matched. Therefore, at the time of creating the composite photo image, it is possible to reduce the error of the subject image shift and create a composite photo image having a subject image very close to the subject.

  According to the third and fourth aspects of the present invention, the X-axis scale and the Y-axis scale as the magnification of the telephoto image before and after conversion are obtained using a circumscribed frame circumscribing the conversion area of the telephoto image reduced by projective conversion. Thus, the projective transformation can be performed without reducing the telephoto image during the projective transformation. Therefore, the reduction in the number of pixels of the telephoto image due to the reduction can be eliminated, and a high-definition image can be displayed superimposed on the wide-angle image.

  Hereinafter, a three-dimensional measurement apparatus according to an embodiment of the present invention will be described with reference to the drawings.

  The three-dimensional measuring apparatus according to the present embodiment measures three-dimensional coordinates of arbitrary measurement points existing in a subject photographed from two or more places. A three-dimensional measuring apparatus is a method for calculating three-dimensional coordinates by a bundle method based on coordinates on a photographic image, and measuring photographic image coordinates (measuring the coordinates of a required reference point on a subject in a photographic image). After the analysis, precise analysis of the photographic image (calculation of the three-dimensional coordinates of the orientation element and reference point by precise analysis of the photographic image based on the approximate value of the three-dimensional coordinates of the orientation element and reference measurement point, as well as any required 3D coordinates of measurement points).

  The orientation elements include three orientation factors such as the two-dimensional coordinates of the photographic principal point position and the screen distance, the distortion coefficient of the photographic image, and the position and tilt of the camera at the time of taking a photograph in an appropriate coordinate system in the subject space. External orientation elements such as a movement amount and three rotation amounts can be given as examples. If the internal orientation element is known, other orientation elements are obtained.

  This apparatus is composed of a computer for inputting a photographic image and performing an analysis operation. The photograph image is preferably taken by a digital camera, but it may be a photograph obtained by digitizing a photograph taken by a film camera with a scanner or the like. The camera may be for photo measurement with internal orientation or a general-purpose digital camera with no internal orientation. A general-purpose computer can be used, and is driven by installing a program that causes the computer to perform processing as described later on the computer using a storage medium that stores the program.

  The photographic image used in this apparatus is most preferably digital information obtained by photographing using a digital still camera or a digital video camera using, for example, a CCD (Charge Coupled Device) as an imaging element. It is also possible to use analog information obtained by photographing with a camera or an analog video camera and converted into digital information. Photo images obtained as digital information can be directly handled by a computer. Note that the resolution, shooting distance, shooting angle, and number of reference measurement points to be shot can be arbitrarily selected according to the need for measurement accuracy and the like. In addition, a hard copy of a photographic image formed on paper, a plastic film or the like can be directly used, and such a hard copy can be used as digital information by an image scanner or the like.

  FIG. 1 shows a block diagram of a photogrammetry apparatus according to an embodiment of the present invention. The photogrammetry apparatus 1 is analyzed by an interface unit 2 for connection with a camera 7 and an input device 8 such as a keyboard and a mouse, a control calculation unit 3 for controlling the entire apparatus, and a display unit 4 for displaying photographic images. A storage section 5 that functions as a storage area and a calculation area for photographs and data, and a system program storage section 6 that functions as a system program by installing the program.

  As will be described later, the control calculation unit 3 is provided with blocks 3a to 3e that are driven by a system program and execute photogrammetry. Each of these blocks functions as a functional block that performs the following processing alone or in cooperation with other blocks in the apparatus.

  Next, the flow of processing when performing photogrammetry using the photogrammetry apparatus of FIG. 1 will be described. FIG. 2 is a flowchart showing the flow of processing when performing photogrammetry using the photogrammetry apparatus of FIG. When photogrammetry is performed using the apparatus 1, the processing of each step shown below is mainly performed by the blocks 3a to 3c of the control calculation unit 3 of the apparatus.

  First, a stereo image is taken, the photographic image is transmitted to the apparatus 1, stored in the image storage unit 5, and the image is displayed on the display unit 4 at the time of operation (# 1). The stereo image is two or more photographic images obtained by photographing the same subject from different positions using the camera 7. Hereinafter, in the present embodiment, when a wide-angle image and a telephoto image taken from the same position are taken as one set, and a total of four photographic images of another wide-angle image and a telephoto image taken from different positions are used. Will be described. A photographic image is distinguished from a set of wide-angle and telephoto images taken from one position as a reference image and the other set as an inspection image. Of the photographic images taken from the two shooting locations, which set of photographic images is used as the reference image is arbitrary, and any one set can be selected.

  In the present embodiment, as described above, in photographing a stereo image, a plurality of photographs having different focal lengths of the camera are photographed at each photographing position. That is, a wide-angle image obtained by photographing a subject with a short focal length and a wide angle, and a telephoto image using a camera with a longer focal length are photographed from the same photographing position (the part of the subject photographed in the wide-angle image). Note that the telephoto image is an image of a part of the wide-angle image, and may be two or more. Further, the telephoto image may be an image taken by changing the direction of the optical axis of the camera, as long as the telephoto image is taken from the same place as the wide-angle image. In the present embodiment, for convenience of understanding, the description will be made assuming that the telephoto images are taken one by one at each photographing position. As will be described later, the telephoto image is used for coordinate measurement when inputting the reference point and the measurement point, and an appropriate portion is taken as the reference point and the measurement point in the wide-angle image. It is preferable. In addition, it is desirable to correct lens distortion in advance for wide-angle images and telephoto images used in this embodiment, but this is not an absolute condition.

  FIG. 3 is a diagram showing a coordinate system for performing photogrammetry. The photographic images 110 and 210 photographed by the two cameras A and B, respectively, have the optical axis direction of the camera as the Z axis and the longitudinal direction of the photographic image as the X axis with respect to the optical axis centers 100a and 200a, respectively. The short direction is the Y axis. The origin is the upper left corner of each photo in the coordinate system on each photo shown below. Therefore, the X axis is not always parallel to the ground surface E according to the tilt of the camera. In the present embodiment, the three-dimensional coordinates are specified based on the camera that captured the reference image. That is, when the photographic image 110 taken by the camera A is used as the reference image, the X1, Y1, and Z1 axes are adopted as the reference coordinate system. Hereinafter, a description will be given of a case where the camera A is a reference camera and a photographic image taken by the camera A is a reference image.

  In the present embodiment, photographic images used for photogrammetry are, for example, as shown in FIGS. 4A, 4B, 5A, and 5B. 4A is a wide-angle image (reference image) taken by camera A, and FIG. 4B is a wide-angle image (inspection image) taken by camera B. 5A shows one of the telephoto images taken by the camera A, and FIG. 5B shows one of the telephoto images taken by the camera B. These photographic images can two-dimensionally represent arbitrary points in the image using the XY coordinate system in units of the photographic images.

  Next, the telephoto image is combined with the wide-angle image (# 2), and the telephoto image is displayed on the display unit 4 in association with the position of the wide-angle image. The composition processing is mainly managed by the image composition unit 3c of the photogrammetry apparatus. FIG. 6 shows a detailed flowchart of the composition process of the wide-angle image and the telephoto image. In the present embodiment, the synthesized image is not a thing in which the telephoto image and the synthesized image are completely merged into one image so that the fineness of the telephoto image is not lost. This means a calculation process used for how to superimpose the images when displayed on the display unit 4.

First, for a wide-angle image and a telephoto image taken from the same position, four or more (four points in the present embodiment) are designated as the composite reference point with respect to the wide-angle image taken from the same location of the subject and the telephoto image. (# 21). This will be specifically described with reference to FIG. FIG. 7 shows a wide-angle image and a telephoto image taken from the photographing point 100a (see FIG. 3). As shown in FIG. 7, for the wide-angle image and the telephoto image, the synthesis reference points 11a to 11d and 12a to 12d indicating the same point of the subject are designated four by four. The combination reference points 11a to 11d and 12a to 12d are designated by operating the input device 8 while referring to the wide-angle image and the telephoto image displayed on the display unit 4. The input points are recognized as the composite reference points 11a to 11d and 12a to 12d by the reference point designating unit 3a. At this time, if a feature portion of the subject such as an end portion of the object shown in the subject is designated as the synthesis reference point, the same point can be easily designated for the wide-angle image and the telephoto image. The coordinates of the composite reference point on the wide-angle image are 11a (x _11a , y _11a ), 11b (x _11b , y _11b ), 11c (x _11c , y _11c ), 11d (x _11d , y _11d ), respectively. The coordinates on the image are 12a ( _x12a , _y12a ), 12b ( _x12b , _y12b ), 12c ( _x12c , _y12c ), 12d ( _x12d , _y12d ), respectively. At this time, in order to acquire a set of designated coordinates, pattern matching can be used for automation.

  By substituting the coordinates of the four sets of composite reference points obtained by associating four or more points into the projective transformation expression shown in Expression (1), nine constants (A, B in Expression (1) are obtained. , C, D, E, F, a, b, c) are calculated (# 22).

式（１）において、ｘ_１２、ｙ_１２は望遠画像側の合成基準点の座標１２ａ〜１２ｄで、ｘ_１１、ｙ_１１は広角画像側の合成基準点の座標１１ａ〜１１ｄを示す。

In Expression (1), x ₁₂ and y ₁₂ are coordinates 12a to 12d of the composite reference point on the telephoto image side, and x ₁₁ and y ₁₁ are coordinates 11a to 11d of the composite reference point on the wide-angle image side.

  By inputting four or more sets of composite reference points, constants (A, B, C, D, E, F, a, b, c) for superimposing the telephoto image on the wide-angle image can be obtained. It is possible to create. However, if this constant is applied as it is and the telephoto image 120 is combined with the wide-angle image 110, the telephoto image 120 is reduced and the number of pixels is reduced, so that the fineness of the image is lost. In addition, when displaying the composite image, a method of displaying the composite image with the telephoto image as an original image while transforming from the above-described constants (1) is conceivable. The display speed of the composite image becomes slow. Therefore, the following processing is performed.

Specifically, as shown in FIG. 8, the coordinates of the outer frame of the telephoto image 120 are converted using the formula (1) from which the above constants are obtained. Specifically, the coordinates of the four corner portions 13a to 13d of the telephoto image 120 may be converted by Expression (1). Points obtained by converting the four corner portions 13a to 13d of the telephoto image 120 according to the expression (1) are referred to as outer frame conversion points 14a to 14d. Here, the coordinates are 14a ( _X14a , _Y14a ), 14b ( _X14b , _Y14b ), 14c ( _X14c , _Y14c ), 14d ( _X14d , _Y14d ) (# 23).

  Next, as shown in FIG. 9, from the coordinates 14a to 14d of the outer frame conversion point, a rectangle circumscribing the region connecting the outer frame conversion points, the horizontal direction along the X axis, and the horizontal direction along the Y axis. The circumscribed frame 15 is calculated (# 24). Specifically, among the coordinates 14a to 14d of the outer frame conversion point, the maximum value and minimum value of the X coordinate and the maximum value and minimum value of the Y coordinate are searched, and 15a (X maximum, Y minimum) and 15b, respectively. The coordinates of four points are determined by a combination of (X maximum, Y maximum), 15c (X minimum, Y maximum), 15d (X minimum, Y minimum), and a range surrounded by the points indicated by the coordinates is defined as a circumscribed frame 15 And it is sufficient. Also, the size Xo (difference between 15a and 15d) of the circumscribing frame 15 and the size Yo (difference between 15c and 15d) of the circumscribing frame 15 are obtained, and X of the entire telephoto image 120 is obtained. Compared with the information of the size Xa in the axial direction (difference between 13a and 13d) and the size Ya in the Y-axis direction of the entire telephoto image 120 (difference between 13c and 13d), X An axis scale Sx and a Y-axis scale Sy are obtained (# 26). Here, the X-axis scale Sx and the Y-axis scale Sy are scales for preventing loss of definition of the telephoto image.

式（２）において、ＳｘはＸ軸スケール、ＳｙはＹ軸スケール、Ｘａは望遠画像全体のＸ軸方向の大きさ、Ｘｏは外接枠のＸ軸方向の大きさ、Ｙａは望遠画像全体のＹ軸方向の大きさ、Ｙｏは外接枠のＹ軸方向の大きさを示す。

In Expression (2), Sx is the X-axis scale, Sy is the Y-axis scale, Xa is the size of the entire telephoto image in the X-axis direction, Xo is the size of the circumscribed frame in the X-axis direction, and Ya is the Y of the entire telephoto image. The size in the axial direction, Yo, indicates the size in the Y-axis direction of the circumscribed frame.

Based on the X-axis scale Sx and the Y-axis scale Sy obtained in the above, the converted telephoto image frame can be calculated by inputting the coordinates 14a to 14d of the outer frame conversion point into the equation (3). Specifically, as shown in FIG. 10, if the coordinates 14a to 14d of the outer frame conversion points are input to the expression (3), the coordinates 16a (x _16a , y _16a of the corner portions of the corresponding converted telephoto image frames are respectively provided. ), 16b ( _x16b , _y16b ), 16c ( _x16c , _y16c ), 16d ( _x16d , _y16d ) can be calculated.

  As described above, since the coordinates of the telephoto image and the coordinates 14a to 14d of the outer frame conversion points have the relationship shown in the equation (1), the definition of the telephoto image is determined by the X-axis scale Sx and the Y-axis scale Sy. Can be converted into a converted telephoto image without impairing the image quality. That is, there is a relationship shown in Expression (4) derived from Expression (1) and Expression (3) between the coordinates of an arbitrary point in the telephoto image and the coordinates of each point in the converted telephoto image.

式（４）において、ｘは望遠画像中の任意の点のＸ座標、ｙは望遠画像中の任意の点のＹ座標、ｘ_１６は変換望遠画像のＸ座標、ｙ_１６は変換望遠画像のＹ座標、ＳｘはＸ軸スケール、ＳｙはＹ軸スケールを示す。

In the formula (4), x is the X coordinate of an arbitrary point in the telephoto image, y is the Y coordinate of an arbitrary point in the telephoto image, x ₁₆ is X coordinate of the transformation telephoto image, the y ₁₆ conversion telephoto image Y Coordinates, Sx indicates an X-axis scale, and Sy indicates a Y-axis scale.

  Therefore, as shown in FIG. 11, the telephoto image 120 is converted into the converted telephoto image 16 based on the equation (4) (# 27). By this processing, projective transformation of an image used when the telephoto image 120 is pasted into a wide-angle image can be performed without losing the definition of the image.

  In addition, in pasting the wide-angle image and the telephoto image, the converted telephoto image is appropriately reduced or enlarged and displayed. Even in this case, the data (number of pixels) of the converted telephoto image is not actually reduced, but is displayed with the same fineness as the telephoto image.

  In addition, the wide-angle image and the telephoto image captured by the camera B are similarly combined, and when this processing ends, the wide-angle image and telephoto image combination processing ends.

  When the composition processing of the wide-angle image and the telephoto image is completed, photographic image coordinate measurement [measurement of required reference point coordinates (two-dimensional coordinates) on the subject in the photographic image] is performed.

  Specifically, as shown in FIG. 2, a reference point is designated in an image obtained by combining a wide-angle image and a telephoto image (# 3). The designation of the reference point is preferably performed in the area of the converted telephoto image displayed by being pasted on the wide-angle image.

  As a reference point, a position that is easily discernable in appearance, such as a corner of a subject, is selected at a portion of the composite image with the telephoto image. Of course, a subject in which a target for use as a reference point in advance is placed on the subject may be photographed. The reference point is designated by inputting an object such as a mouse to a portion of a subject that is commonly displayed in a region where a telephoto image is pasted in two composite images taken from two shooting locations. Input by the operation. The input point is recognized as a reference point by the coordinate point designating unit 3a.

  These processes are repeated to select five or more reference points in both images, and the positional relationship between the camera A100 and the camera B200 is calculated using the coordinate values of each reference point on the photographic image (# 4). This embodiment is called external orientation in the photogrammetry method, and as a positional relationship, the optical axis center of the camera B200 (inspection camera) is the reference coordinate system having the optical axis center of the camera A100 (reference camera) as the origin. A coordinate value indicating where it is located and an angle indicating in which direction the optical axis of the camera B200 is directed are derived.

  The position calculation unit 3b of the device 1 controls the calculation of the photographing position relationship. In the present embodiment, as a shooting position relationship, another camera, that is, a camera at a shooting position B (hereinafter referred to as a camera) as viewed from a coordinate system based on a camera at the shooting position A (hereinafter abbreviated as camera A). Abbreviated as B.) and coordinates indicating where the camera B is located and an angle indicating which direction the camera B is facing.

  Note that, in the calculation of the photographing position relationship, the coordinates of the reference point are performed using coordinate values that are finely defined in the telephoto image. That is, since the telephoto image in the composite image includes finer image information than the wide-angle image, it is possible to specify position information with higher accuracy by using the coordinate value specified by the telephoto image. It is possible to improve the accuracy of calculation of the photographing position relationship of the camera. At this time, arbitrary coordinate values in the converted telephoto image are converted into coordinate values on the wide-angle image by the following expression. The converted coordinate value on the wide-angle image is calculated as a value after the decimal point using the floating point function stored in the control calculation unit 3 of the photogrammetry apparatus (see FIG. 12).

式（５）において、ｘ₁₀は広角画像中のＸ座標、ｙ₁₀は広角画像中のＹ座標、ｘ_１６は変換望遠画像のＸ座標、ｙ_１６は変換望遠画像のＹ座標、ＳｘはＸ軸スケール、ＳｙはＹ軸スケール、ｘ_０は変換望遠画像の原点のＸ座標が広角画像中に対応する点のＸ座標、ｙ_０は変換望遠画像の原点のＹ座標が広角画像中に対応する点のＹ座標を示す。

In Expression (5), x ₁₀ is the X coordinate in the wide-angle image, y ₁₀ is the Y coordinate in the wide-angle image, x ₁₆ is the X coordinate of the converted telephoto image, y ₁₆ is the Y coordinate of the converted telephoto image, and Sx is the X axis. scale, Sy is the Y-axis scales, x ₀ is the X coordinate of the point X coordinate of the origin of the conversion telephoto image corresponds in wide-angle image, y ₀ is the point where the Y coordinate of the origin of transformation telephoto image corresponds in wide-angle image Indicates the Y coordinate.

  When the shooting position relationship of the camera is specified, next, a measurement point in the subject as a measurement target is designated. The measurement point is preferably specified in the area of the converted telephoto image displayed on the wide-angle image.

  For specifying the measurement point, a portion of the subject that is shown in common in the area where the telephoto image is pasted and displayed in the reference image is input by operating the input device 8 such as a mouse. The input point is recognized as a measurement point by the coordinate point designating unit 3a.

  When the input of the measurement point is completed, the position calculation unit 3b analyzes the three-dimensional coordinates in the subject three-dimensional system of the measurement point. The position analysis of the measurement point is preferably performed by the least square method using the bundle method. At that time, necessary repetitive calculations are performed. In addition, in the analysis by the least square method using the bundle method, as described above, the coordinate values in the photograph used for the position analysis of the measurement points are performed using the coordinate values made fine in the telephoto image. . That is, since the telephoto image in the composite image includes finer image information than the wide-angle image, it is possible to specify position information with higher accuracy by using the coordinate value specified by the telephoto image. it can.

  For ease of measurement work and improvement of analysis accuracy, the self-calibration method can be used to calculate the internal orientation elements, so that the internal orientation elements at various sites are unknown or sufficiently verified. Measurement work using a camera that is not available can be facilitated.

  As described above, according to the three-dimensional measurement apparatus according to the present embodiment, a telephoto image and a wide-angle image are used so that the subject image in the telephoto image matches the subject image in the wide-angle image using the telephoto image and the wide-angle image. In the display of the composite photographic image, a shooting area in which the telephoto image is combined using a composite image that displays the telephoto image without losing the fineness of the image. Since the same point of the subject can be selected with high accuracy by selecting the reference point, the position information can be calculated and the three-dimensional position information of the measurement point in the subject space can be measured with high accuracy.

  In addition, since the composition of the wide-angle image and the telephoto image is performed using projective transformation, the optical axes of the two are greatly shifted, for example, when the telephoto image exists in the end region of the wide-angle image. Even if the image is taken, it is possible to convert the telephoto image as if it was taken from the optical axis of the wide-angle image, and to match the subject images of the two photographic images, so that the composite photographic image does not cause a deviation of the subject images. Can be created. Further, by using the X-axis scale and the Y-axis scale, it is possible to eliminate the reduction of the telephoto image that occurs at the time of projective transformation, and to eliminate the reduction in the number of pixels of the telephoto image displayed in the wide-angle image, thereby obtaining a high-definition composite photograph image. Can do.

  In addition, this invention is not limited to the said embodiment, It can implement in another various aspect.

  The telephoto image does not have to be one for each wide-angle image, and may be composed of a plurality of images obtained by photographing a plurality of regions of the wide-angle image. In this case, it is preferable that the projection conversion of the telephoto image is repeatedly performed to generate a composite photo image in which all the telephoto images are combined and displayed, and then the reference point designation process is performed.

It is a block diagram of the photogrammetry apparatus which concerns on embodiment of this invention. It is a flowchart which shows the flow of the whole process in the case of performing photogrammetry using the photogrammetry apparatus of FIG. It is a figure which shows the coordinate system about the case where photogrammetry is performed. It is an example of a wide-angle image photographed by camera A. It is an example of a wide-angle image photographed by the camera B. It is an example of a telephoto image image | photographed with the camera A. It is an example of the telephoto image image | photographed with the camera B. It is a detailed flowchart of the process which synthesize | combines by the projection transformation about the imaged wide-angle image and a telephoto image. It is explanatory drawing of the process in selection of the synthetic | combination reference point of the image of 2 sheets synthesize | combined. It is explanatory drawing of the process which carries out projective transformation of the outer frame of a telephoto image, and produces an outer frame conversion point. It is explanatory drawing of the process which calculates an X-axis scale and a Y-axis scale based on the circumscribed frame circumscribing an outer frame conversion point. It is a conceptual diagram about a conversion telephoto image. It is explanatory drawing about the process which produces a conversion telephoto image from a telephoto image. It is a figure explaining conversion of the coordinate value of the arbitrary coordinates in a conversion telephoto image, and the arbitrary images with a wide angle image.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Photogrammetry apparatus 2 Interface part 3 Control calculating part 4 Display part 5 Image memory | storage part 6 System program 7 Digital camera 8 Input device 11a-11d Wide angle side synthetic | combination reference point 12a-12d Telephoto side synthetic | combination reference point 13 Corner part 14 of a telephoto image Outer frame 14a to 14d Outer frame conversion point coordinate 15 Outer frame 15a to 15d Coordinate of corner part of outer frame 16 Converted telephoto image 100 Camera A (reference camera)
100a Optical center 110 of camera A Wide-angle image 120 photographed by camera A Telephoto image 200 photographed by camera A Camera B (inspection camera)
200a Optical center of camera B 210 Wide-angle image taken by camera B 220 Telephoto image taken by camera B

Claims (5)

Using a plurality of photographic images obtained by photographing a subject by a camera disposed at a reference position and a camera disposed at an inspection position different from the reference position, a reference point that is commonly displayed in the plurality of photographic images. A photographing position relationship between the reference position and the inspection position is calculated based on the coordinate value in the photographic image, and the measurement point is calculated from the coordinate value of an arbitrary measurement point in the photographic image based on the information on the photographing position relationship. A 3D measurement method for measuring 3D position information in a subject space,
As the photographic images photographed from the reference position and the inspection position, respectively, a wide-angle image photographed with a short focal length, and a focal distance obtained by photographing a part of the photographing region from the same photographing position as the wide-angle image are obtained. Using at least one telephoto image taken in a long state, a composite photo image obtained by combining and displaying the telephoto image in the wide-angle image taken from the same photographing position is created for each of the photographing positions, A composite photo image,
In the composite photo image created for each shooting position, select a reference point that is commonly shown in the shooting area,
A three-dimensional measurement method characterized by calculating a photographing position relationship between the reference position and the inspection position based on coordinate values in the plurality of composite photograph images of the reference point. The composite photograph image selects four or more reference points that are common to the wide-angle image and the telephoto image taken from the same photographing position,
Based on the coordinate value of the reference point in the wide-angle image and the telephoto image, a projective transformation formula is calculated, and the telephoto image is generated and synthesized and displayed in the wide-angle image by the projective transformation formula. The three-dimensional measurement method according to claim 1.   The composite image includes an X-axis direction dimension of a circumscribed frame circumscribing a conversion region obtained by projectively transforming the telephoto image into the wide-angle image based on the conversion formula, an X-axis direction dimension of the telephoto image having a Y-axis direction dimension, and The composite telephoto image is created by combining and displaying the converted telephoto image in the wide-angle image based on an X-axis scale and a Y-axis scale respectively derived as a ratio to a dimension in the Y-axis direction and a projective transformation formula. Item 3. A three-dimensional measurement method according to item 2. 4. The three-dimensional measurement method according to claim 3, wherein the converted telephoto image is created by converting each point of the telephoto image according to the following equation.

In the above formula, x is the X coordinate of an arbitrary point in the telephoto image, y is the Y coordinate of an arbitrary point in the telephoto image, x ₁₆ is X coordinate of the transformation telephoto image, the Y coordinate of y ₁₆ converts telephoto image, Sx is an X-axis scale, Sy is a Y-axis scale, A, B, C, D, E, F, a, b, and c are constants of the projective transformation formula. Using a plurality of photographic images obtained by photographing a subject by a camera disposed at a reference position and a camera disposed at an inspection position different from the reference position, a reference point that is commonly displayed in the plurality of photographic images. A photographing position relationship between the reference position and the inspection position is calculated based on the coordinate value in the photographic image, and the measurement point is calculated from the coordinate value of an arbitrary measurement point in the photographic image based on the information on the photographing position relationship. A three-dimensional measuring apparatus that measures three-dimensional position information in a subject space,
As the photographic images photographed from the reference position and the inspection position, respectively, a wide-angle image photographed with a short focal length, and a focal distance obtained by photographing a part of the photographing region from the same photographing position as the wide-angle image are obtained. Image input means for inputting photographic image data with at least one telephoto image taken in a long state;
Coordinate input means for inputting a coordinate value on the photographic image for an arbitrary point of the photographic image input by the image input means;
Image composition means for creating, for each photographing position, a composite photograph image obtained by compositing the telephoto image in the wide-angle image photographed from the same photographing position among the photographic image data input by the image input means;
In the composite photographic image created for each shooting position, the coordinate value input by the coordinate input means for the same position of the subject that is common in the shooting area, as the coordinate value of the reference point, Based on the coordinate value, a positional relationship calculating means for calculating the shooting positional relationship of the photographic image;
With respect to the measurement position of the subject in the composite photograph image, the coordinate value input by the coordinate input means is used as the coordinate value of the measurement point, and based on the information on the photographing position relation calculated by the positional relation calculation means. And a measuring point calculation means for calculating the three-dimensional coordinates in the subject of the measuring point reflected in the composite photograph image.

Images