JP2002236021A - Coordinate calculation method of survey point, matching program and land mark in precision photo survey - Google Patents

Abstract

PROBLEM TO BE SOLVED: To easily and precisely acquire the photo survey coordinates of land marks as surveying points to make data for subsequent analysis quickly preparable. SOLUTION: In the calculation of the coordinates of the land marks (1 to 5) projected on images 1 and 2 of two digital photographs respectively, a non-linear analysis is made while applying a collinear conditional expression of a single photograph to the images 1 and 2 photographed so as to include the land marks (1 to 5), initial approximate value information at each photographing position as analysis information of the conditional expression, the coordinates of a reference point and initial approximate value coordinates are inputted, with respect to information on the two image taken into a computer. Photographic coordinates of the reference value is obtained for each of the called images and a specified fixed value is substituted for internal localization elements under inputted respective conditions to solve the collinear conditional expression, thereby accomplishing a matching between the land marks within the two images. Thereafter, the non-linear analysis if carried out using the coordinate information subjected to the matching as initial value to calculate the coordinates of the individual land marks.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for calculating coordinates of a survey target point in precision photogrammetry, a matching program, and a landmark, and can easily and accurately acquire a photogrammetric coordinate of a landmark as a survey target point. The present invention relates to a method for calculating coordinates of a survey target point, a matching program, and a landmark, which can quickly generate data for the analysis.

[0002]

2. Description of the Related Art In recent years, in the field of photogrammetry, an image of a digital still photograph taken by a digital camera is processed using a personal computer, and the three-dimensional coordinates of the photographed object are precisely obtained. A precision photogrammetry technique capable of performing calculations, calculation of target surface shapes, dimensions, and the like has been developed. In this type of precise photogrammetry, a survey object is photographed from multiple angles with a digital camera, and the image data is subjected to orientation analysis of the image using image processing software of a personal computer, etc., and the photographed object or object is photographed. Accurate coordinates of feature points (gauge points) (hereinafter, referred to as landmarks) of various types of surveying objects installed for specifying a part of an object are acquired.

[0003] In this regard, as a precision photogrammetry method described above, the applicant has applied a non-linear collinear condition formula, which is established between a plurality of overlapping photographs taken by a digital camera, to Self C.
We are developing a system that can obtain high-precision three-dimensional absolute coordinate values and specify the surface of the target part by solving the alibration method and obtaining accurate photo coordinates. Hereinafter, the analysis by the Self Calibration method will be briefly described. FIG. 8 is a schematic explanatory diagram showing the collinear condition of a plurality of single photographs solved using the Self Calibration method. In the figure, the surveying object point P _M in the model space, collinear conditions with a point on the two photographic images taken at two points (Picture1,2) is shown. At this time, the three-dimensional coordinates (X _j ,
The photographic coordinates (x _ij , y _ij ) of the point j in the photographic image i such as Y _j , Z _j ) are represented by the following equations (1) and (2).

[Equation 1] Here, the photograph i has six external orientation elements (X _0i , Y _0i , Z
_0i : camera projection center coordinates, ω _i , φ _i , κ _i : camera tilt). Also, correct photo coordinates (x _ij , y _ij )
Is the photo coordinates (x _cij , y) acquired on the image at the time of analysis
_cij ) is expressed by the following equations (4) to (8).

[Equation 2] Here, as internal orientation elements, c _i : focal length of the camera at the time of photograph i, x _hi , y _hi : displacement of principal point position, k _1i , k
_2i , _k3i , _p1i , _p2i : coefficients of distortion of camera lens

The correct photographic coordinates (x _ij , y _ij ) are obtained by performing a predetermined iterative calculation to solve the nonlinear collinear conditional expressions represented by the expressions (1) to (8) based on the above conditions. Can be requested.

FIG. 9 is a block diagram showing a series of processing steps of a precision photogrammetry system incorporating the above-described analysis method. As shown in the figure, first, a landmark is set at a position to be a surveying target point among topography and a structure as a surveying target object (step 500). in this case,
A landmark may be used as a landmark without setting a landmark. Next, a plurality of images are photographed by changing the photographing position with respect to the survey target using a digital camera having specifications suitable for surveying (step 510). At this time, information data of the photographing position is collected. The photographed image data is input to the personal computer via the data medium of the digital camera (step 520). Analysis data for coordinate calculation is created from the photograph coordinates of each image on a personal computer (step 530), the above-described nonlinear analysis is performed, three-dimensional coordinate calculation of landmarks, strike of a surface specified by a plurality of landmarks is performed. , Inclination and the like can be calculated (step 540). The analysis result is
A list can be output to a known spreadsheet program or an appropriate output diagram can be output by a plotting program such as a post processor (step 550). Further, the data can be transmitted to another process and used as needed (step 560).

In conventional analytical photogrammetry, a plurality of landmarks are arranged on an object to be photographed, two pictures are taken at slightly different photographing angles, and an analytical plotter such as a stereo comparator is used. In addition, the coordinates (photo coordinates) of the target landmark are manually obtained from the two photographs, and the correspondence between the corresponding landmarks between the photographs is also manually performed.

[0009] Since these operations need to be performed while checking the image, it is complicated and time-consuming, and the matching (matching) of landmarks between a plurality of photos depends on the obtained photos. In some cases, it is difficult to determine the depth due to the installation position, and it may not be possible to determine the distance position. In addition, it is necessary to give congruent conditions between the photograph coordinates and the actual absolute coordinate system. Conventionally, staffs (staffs) are placed and imprinted on photographs, or between landmarks installed at appropriate intervals. Although the baseline distance has been measured, there are problems that the accuracy of information in the depth direction cannot be sufficiently obtained, and that the baseline distance must be measured.

The accuracy of the absolute coordinates of the landmark finally obtained also differs depending on whether the photographic coordinates of the landmark imprinted on the image taken by the digital camera can be acquired with high accuracy. For this reason, it is preferable to adopt landmarks that allow the installation points to be clearly identified on the photographic image when the target terrain or structure is photographed with a digital camera.

On the other hand, since the above-described nonlinear analysis of the collinear condition expression can be performed on a personal computer, the correspondence between landmarks between photographs, that is, matching is performed on the same personal computer as a preprocessing operation for the analysis. If this is possible, the value can be directly used as an initial approximation value for nonlinear iterative calculation, which contributes to the efficiency of analysis.

[0012] Accordingly, an object of the present invention is to solve the above-mentioned problems of the prior art and to obtain the coordinates of landmarks imprinted on a plurality of photographs taken by a digital camera with high accuracy. It is an object of the present invention to provide a method and a landmark for matching a reference point in precision photogrammetry, which can easily perform landmark matching between photographs.

[0013]

In order to achieve the above object, the present invention provides a single photo collinear condition formula for a plurality of digital photo images obtained by photographing a predetermined survey target area. In the method for calculating the coordinates of the survey target point in precision photogrammetry, which performs the nonlinear analysis by applying and calculates the coordinates of the survey target point projected on the image, the plurality of images captured on the computer are called. The initial approximate value information of the photographing position as the analysis information of the collinear conditional expression, the reference point coordinates of the survey target points, and the initial approximate value coordinates of the survey target points are input, and the called is performed. The photograph coordinates of the reference value are obtained for each of a plurality of images, and a predetermined fixed value is substituted into an internal orientation element based on each of the input conditions to solve the collinear conditional expression, thereby obtaining the plurality of images. Each survey object in the image of Matching is performed between, characterized in that to perform the coordinate calculation of the surveying target point by performing the non-linear analysis based on the coordinate information which is the matching.

[0014] At this time, it is preferable that the center of gravity of the image of the survey target point is acquired as the photograph coordinates of the survey target point from the color tone of the survey target point which is imprinted on the image.

Further, landmarks are set as the survey target points, a plurality of digital photographs of the survey target area including the landmarks are taken, and the coordinates of the photograph are calculated based on the shape of each landmark imprinted on the image. Is preferably obtained.

In order to carry out the above-mentioned method, the landmark is characterized in that a reflecting material is arranged on a surface directly facing the photographing camera position.

At this time, it is preferable that the reflecting material is arranged on one surface of the block.

Further, as a program for performing the above-mentioned matching, the plurality of images captured on a computer are displayed on a screen, and an input unit analyzes the collinear condition expression analysis information input as the analysis condition from the screen. Initial approximate value information of the photographing position as, reference point coordinates of the survey target points, initial approximate value coordinates of the survey target points, and a photograph of the reference value for each of the plurality of photographs by an image processing unit The collinear line in which a predetermined fixed value is assigned as a default initial value to an internal orientation element by an arithmetic unit based on the obtained photograph coordinates obtained by calculating the image centroid displayed on the screen using the coordinates as pixel information. The computer is caused to execute a procedure of repeatedly calculating a conditional expression, performing matching between the respective survey target points in the plurality of images, and obtaining corresponding photograph coordinates and corresponding numbers between the respective survey target points. And wherein the door.

[0019]

BEST MODE FOR CARRYING OUT THE INVENTION
Performed in the method of calculating the coordinates of the survey target point
Of landmarks suitable as a surveying method and a survey target point
Embodiments will be described with reference to the accompanying drawings. [Matching of landmarks-calculation of coordinates of survey target points]
In the present embodiment, five landmarks as shown in FIG.
In the survey target where a mark (marker) is installed, the land
Based on an example where the three-dimensional coordinates of both ends of the mark are measured
Next, the procedure of the matching method will be described. Ma
First, the five landmarks 1 shown in FIG.
~ 5 (X_j, Y_j, Z _jTerrain including j = 1-5)
Shoot with a tall camera. Shooting at two shooting positions
An example of the images 1 and 2 will be described. Approximate shooting at this time
Distance between the shadow position and the reference point (shooting distance), digital data to be used
Camera specifications, approximate shooting information (focal length, shooting
Approximate tilt of camera with respect to target (κ, φ, ω)
Collect. In addition, in FIG.
The landmarks 1 and 5 at the ends are used as reference points.

The following processing and analysis using a program for the functions of a matching method (described later) and a collinear conditional expression analysis method provided in the personal computer will be described. First, target image data is taken from a recording medium of a digital camera into a personal computer. In this case, image files of various formats supported by image processing software used as a preprocessor can be handled. The following operation will be described with reference to FIG.

First, a plurality of image files taken into a personal computer are opened (step 100). At this time, it is preferable that the black and white display of the image is inverted so that the landmark is displayed as a black point on a white background in order to clearly recognize a landmark described later. Furthermore, as information of the image file opened on the desktop,
Number of landmarks, number of reference points, number of length information imprinted,
The number of repetition calculations is input to the displayed predetermined input window (step 110).

Next, among the surveying objects, the reference point coordinates (X _i , Y _i , Z _i : unit mm) whose coordinates are known are determined by the number of reference points (i = 1 to 1).
2) Input minutes (step 120). In the present embodiment, the coordinates of the reference point are distances from the origin in a predetermined absolute coordinate system.

On the other hand, as the information on the photographing position, the photographing distance from the camera to the reference point and the focal length of the camera at the time of photographing are input as fixed values for the initial calculation (step 13).
0). In addition, by inputting the image scale, the distance (pixel length) indicated by one pixel on the image from the reference point coordinates can be obtained as scale information.

Further, the photograph coordinates of the reference point on the image are obtained (step 140). At this time, the center of gravity of the pixel group displaying each reference point can be calculated and obtained as the photograph coordinates of each reference point.

At this time, an initial approximate value of the photographing position and the angle of the camera (photographing angle) is input as data on the situation in which each photograph was taken (step 150). Hereinafter, analysis using the above-described collinear condition expression is performed, and matching of a predetermined landmark is performed. Here, for the initial calculation, the displacement of the principal point of the lens and the coefficient of the distortion of the camera lens are set as fixed values (zero).

After inputting the initial setting values, the approximate value of the three-dimensional coordinates of the target landmark is input as an initial approximate value (step 160). The three-dimensional coordinates at this time are different from the initial values of the subsequent iterative calculations, and may be values that serve as a guide for the matching method. Here, by performing a tentative calculation based on the input initial approximation value, provisional coordinates of three-dimensional coordinates serving as a guide can be calculated and superimposed on the photographed image, thereby making the input value valid. Can be confirmed.

If it is confirmed on the screen that there is no large displacement between the provisional coordinates and the landmark image, a search is made for a landmark on the image near the provisional coordinates, and the photographic coordinates of the landmark are obtained (step). 170). Further, length information of the staff projected on the image is input, a landmark number constituting a vertex of the surface to be calculated is input, and the surface is specified (steps 180 and 190).

By inputting the coordinates of the reference point and the coordinates of each landmark through the above-described processing procedure, the above-described nonlinear collinear condition expression is converted into the internal orientation elements (focal length of camera, displacement of principal point position, camera lens The coefficient can be matched by applying the Self Calibration method with a fixed value of the coefficient of distortion. Then, by applying the Self Calibration method based on the coordinate values of the known reference points, the camera position information as the external orientation element initially given is changed, and the three-dimensional coordinate values of all landmarks are obtained. Can be.

[Modification of Matching Method] Instead of obtaining the above-described photo coordinates, all the photo coordinates are obtained only from the first image, and the three coordinates of each landmark of the other image are calculated from the coordinates. Dimensional coordinates can be obtained. That is, in the first embodiment, all three-dimensional coordinates of a landmark are obtained and input, but the landmark can be used as a calculated value in other images based on the result.

[Regarding Landmarks] When photographing the above-mentioned surveying object, landmarks are set at reference points and singular points of the terrain so that the landmarks are clearly reflected in the image. It is preferable that the photograph coordinates can be obtained with high accuracy. For this reason, landmarks having the configuration shown in FIGS. 3 to 7 can be considered as landmarks used for photographing. FIG. 3A includes a reflector plate (reflection plate) 10 having the same configuration as that of a target for optical wave measurement, and a high-intensity reflector 1 is adhered to the surface thereof.
The reflector plate 10 is used so as to face the camera by leaning against a structure to be surveyed or a pile provided on the ground surface. FIG. 2B shows a cubic block 11 on one side of which a reflecting material 1 is attached, in which the stability at the time of installation is considered in comparison with the reflector plate 10. FIG. 3C shows a spherical landmark 12. Since the spherical surface of the spherical landmark is formed of the reflecting material 1, the center-of-gravity coordinate value can be obtained with high accuracy when centering on an image. In addition, by using a spherical light bulb having a similar shape, a self-luminous landmark can be obtained.
They are very useful for night surveys and photogrammetry in tunnels. FIG. 3D shows a landmark 13 also serving as a target for the optical wave measurement, and a mark 2 for collimation is written at the center of the landmark 13. Also in this case, by finishing the entire surface with the reflective material 1, the visibility of the object at the time of light wave distance measurement and photographing by the digital camera can be improved.

In the above description, the visibility at the time of photographing is enhanced by using a reflective material on the surface facing the camera.
It is also preferred that landmarks are easily identified on an image by arranging them with a certain regularity.

FIG. 4 shows an example of a scale 20 used for imprinting accurate length information on a photograph. The scale 20 has arms 22 extending in two orthogonal directions at the bottom of the main body pole 21. By extending each arm 22, it can be used as an index of a reference length in three orthogonal directions. A reflective mark 23 in the form of a seal is attached to the pole 21 of the scale 20 and the end of each arm 22 shown in FIG.
A reference length (for example, L _x , L _{y in the} figure) is defined between the marks 23. In addition, in order to install the length information scale on a steep slope, the pole may be a pile type and an orthogonal arm may be attached to the pole upper end. In addition, in a place such as a cliff, a launching device may be used to fly a sagittal scale from a distant place and land on a part of the cliff.

FIGS. 5A and 5B show examples of the installation of landmarks which can be adjusted in angle so that the reflection plate faces the camera. In FIG. 5A, a reflector 30 is attached to a support pole 31 driven into an inclined surface via an angle adjusting bracket 32. By adjusting the tilt angle of the angle adjusting bracket 32, the reflecting plate 30 can be moved according to the gradient by a camera (not shown).
Can be confronted. FIGS. 7B and 7C show an example in which landmarks are installed by using a cover cap 5 of an anchor head of a ground anchor that is cast on a slope. FIG. 3B shows a structure in which the reflection plate 30 is attached to the upper end of the cover cap 5 via an adhesive 33 made of resin clay or the like attached to the back side of the reflection plate 30. When the adhesive 33 of the reflection plate 30 is pressed against the upper end of the cover cap 5 so that the reflection plate 30 maintains an appropriate angle, the adhesive 33 is deformed, and after a predetermined time elapses, the adhesive 33 is solidified in that state. Can be fixed to the upper end of the cover cap 5. FIG. 3C shows a structure in which the reflection plate 30 is attached to the upper end of the cover cap 5 via the attachment fitting 34. The mounting bracket 34 is mounted on the upper end of the cover cap via a fixing screw 35 screwed to a ring-shaped mounting frame 36. Mounting bracket 3
The reflector 4 is attached to the cover 4 so that the angle can be adjusted. The angle of the reflector 30 may be adjusted after the mounting frame 36 is attached to the upper end of the cover cap 5.

FIG. 6 shows a landmark in which a reflector 40 is attached to a mesh intersection of a mesh 41 formed by knitting a synthetic fiber rope corresponding to the size of a target structure. According to the landmark having such a configuration, as shown in FIG. 7, by laying the mesh 41 in accordance with the change of the terrain, the reflector 40 is arranged following the terrain, and the characteristics of the terrain can be accurately determined. Can catch well.

In addition, in the case of a steep terrain where landmarks cannot be set up on site, a so-called color ball containing a liquid coloring material inside is hit against the object to be measured to remove the coloring material in the color ball. By scattering, a part of the terrain can be colored to become a landmark.

[0036]

As described above, the work of matching landmarks between a plurality of photographs, which is performed in precision photogrammetry, can be performed accurately and promptly, whereby the three-dimensional coordinates of the survey target point can be obtained. This can be effectively used also as information for analysis for obtaining. Also,
There is also an effect that the accuracy of data at the time of analysis can be increased by the used landmark.

[Brief description of the drawings]

FIG. 1 is an explanatory diagram schematically showing a positional relationship between an image and target coordinates when calculating coordinates of a survey target point in precision photogrammetry according to the present invention.

FIG. 2 is a flowchart showing a procedure of a method of matching a reference point in precision photogrammetry according to the present invention.

FIG. 3 is a perspective view showing an embodiment of a landmark as a survey target point.

FIG. 4 is a perspective view showing an embodiment of a scale as a survey target point.

FIG. 5 is a side view showing an embodiment of a landmark as a survey target point.

FIG. 6 is a partial perspective view showing a modified example of a landmark as a survey target point.

FIG. 7 is an explanatory view schematically showing a usage example of the landmark shown in FIG. 6;

FIG. 8 is an explanatory view schematically showing a positional relationship between an image and target coordinates when calculating coordinates of a survey target point in conventional precision photogrammetry.

FIG. 9 is a flowchart showing an example of a conventional procedure for calculating and analyzing coordinates of a survey target point in precision photogrammetry.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Reflector 10, 30, 40 Reflector 11 Block 20 Scale 23 Mark

Continuing on the front page (72) Inventor Hirotaka Nakahara No. 2, Sanbancho, Chiyoda-ku, Tokyo Tobishima Construction Co., Ltd. (72) Inventor Kazunobu Matsumoto No. 2, Sanbancho, Chiyoda-ku, Tokyo Tobishima Construction Co., Ltd. (72 ) Inventor Masayuki Tsutsui 2nd Sanbancho, Chiyoda-ku, Tokyo Inside Tobishima Construction Co., Ltd. (72) Inventor Yuki Kumagai 2nd Sanbancho, Chiyoda-ku, Tokyo Tobishima Construction Co., Ltd.F-term (reference) BB28 BB29 DD06 EE00 FF05 FF09 FF61 JJ03 JJ05 JJ26 QQ03 QQ17 QQ24 QQ25 QQ26 QQ27 QQ28

Claims (6)

[Claims] 1. A non-linear analysis is performed on a plurality of digital photograph images obtained by photographing a predetermined survey target area by applying a collinear conditional expression of a single photograph, and the image is projected on the image. In the method of calculating the coordinates of a survey target point in precision photogrammetry for calculating the coordinates of the survey target point, the plurality of images captured on a computer are called, and an initial photographing position as analysis information of the collinear conditional expression is used. Approximate value information, reference point coordinates of the survey target points, and initial approximate value coordinates of the survey target points are input, and photograph coordinates of the reference values are acquired for each of the plurality of called images. Based on each of the input conditions, a predetermined fixed value is substituted into an internal orientation element to solve the collinear condition equation, and perform matching between the respective survey target points in the plurality of images, Matched coordinate information A method of calculating coordinates of a survey target point in precise photogrammetry, wherein the coordinates of the survey target point are calculated by performing the non-linear analysis based on 2. The method according to claim 1, wherein an image center of gravity of the survey target point is obtained as a photograph coordinate of the survey target point from a color tone of the survey target point projected on the image.
3. The method for calculating coordinates of a survey target point in precision photogrammetry described in 4. above. 3. A landmark is set as the surveying target point, a plurality of digital photographs of a surveying target area including the landmark are taken, and a photograph coordinate is calculated from the shape of each landmark imprinted on the image. 3. The method of calculating coordinates of a survey target point in precision photogrammetry according to claim 2, wherein 4. The landmark according to claim 3, wherein a reflecting material is provided on a surface facing the photographing camera position. 5. The landmark according to claim 4, wherein said reflecting material is arranged on one surface of the block. 6. The collinear conditional expression according to claim 1, wherein the plurality of images captured on the computer are displayed on a screen, and input as analysis conditions from the screen by the input unit. Initial approximate value information of a photographing position, reference point coordinates of the survey target points, initial approximate value coordinates of the survey target points, and a photograph coordinate of the reference value for each of the plurality of photographs by an image processing unit. The collinear conditional expression in which a predetermined fixed value is assigned as a default initial value to an internal orientation element by an arithmetic unit based on the obtained photograph coordinates obtained by calculating the image centroid displayed on the screen as pixel information. And performing a process of repeatedly calculating and performing matching between the respective survey target points in the plurality of images to obtain the corresponding photograph coordinates and corresponding numbers between the respective survey target points. For precision photogrammetry Program for surveying target points.