JPH10318715A - Image measuring apparatus - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain an image measuring apparatus by which a locating, operation and a measuring operation are performed with good efficiency and with high accuracy by a method wherein, on the basis of a positional relationship which is found by an approximate-position measuring part, a reference data block is set at a reference image and at least one out of the setting of a search region and the setting of a search data block in a search image is performed. SOLUTION: An image measuring apparatus 30 is provided with an approximate- position measuring part 31, with a data setting part 32 and with a correspondence discrimination part 33. The approximate-position measuring part 31 uses a feature pattern image whose positional relationship is known in advance, and it approximately finds the positional relationship of an object, to be measured, in one pair of images from the images which are obtained by photographing the object, to be measured, from different directions. In the data setting part 32, one out of the pair of images is set as a reference image, and the other is set as a search image. On the basis of the positional relationship which is found by the measuring part 31, a reference data block is set at the reference image, and a search region is set at the search image. The correspondence discrimination part 33 finds the correspondence relationship between a search data block to be set in the search region and the reference data block, and a three-dimensional measuring operation is automated.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION Automated orientation work prior to three-dimensional measurement from different camera positions, and
The present invention relates to a technique for automatically acquiring an initial value of stereo matching and automating three-dimensional measurement.

[0002]

2. Description of the Related Art In terrestrial photogrammetry, when performing three-dimensional measurement, measurement is performed according to the flow shown in FIG. That is, stereo shooting of the object (a) → position detection (b) → location (c) → stereo model formation (d) → three-dimensional measurement (e)
Is required. These processes are mainly performed by a computer, in which the position detection (b) and the three-dimensional measurement (e) are conventionally performed manually. The position detection (b) is a pre-orientation process for obtaining the position, inclination, and the like of the camera to be photographed.

By obtaining the positional relationship between the cameras and the object to be measured in the orientation (c), a stereo model that can be viewed stereoscopically can be formed, and three-dimensional measurement can be performed. The processing of position detection (b) for performing the orientation (c) is an operation of obtaining coordinate positions on each camera of six or more corresponding points projected on different cameras.

There are two types of three-dimensional measurement (e): point measurement and surface measurement.

[0005] In the case of point measurement, measurement points on an object are manually measured manually, but in order to achieve automation and improve accuracy, a work of attaching a mark to the object is performed. I have.

In the case of surface measurement, it is automatically performed using a technique called stereo matching by image processing. At that time,
Initial settings such as determination of a template image and setting of a search width have to be performed manually.

[0007]

In the operation of position detection (b), the operator selects six or more measurement points on the object in the photographed left and right images, and observes each image to check the position on the object. And the coordinate position detection. However, since these operations basically need to be performed while viewing in a stereoscopic manner, it requires skill and is complicated, difficult, and difficult.

In particular, in the work of determining measurement points, associating left and right images, and detecting detailed position coordinates, individual differences are likely to occur, and it is difficult for each person to obtain sufficient accuracy. In addition, there were many cases where measurement was impossible for some people. In order to avoid such inconveniences, processing such as attaching a mark to an object may be performed.

[0009] However, increasing the number of operations for attaching a mark to the object itself is a minus, and since some objects cannot be easily marked, this method is not widely used.

On the other hand, there is also a method in which two cameras are firmly fixed on a stereo gantry to perform three-dimensional measurement without performing orientation work.

However, in such a case, the positional relationship between the two cameras on the carriage must not be changed, and the measurement environment and the object to be measured are greatly limited. At the same time, such devices are large, heavy, difficult to carry, and expensive, so that three-dimensional measurement by this method is not widely used.

In the point measurement of the three-dimensional measurement (e), when a mark or the like is not put on an object to be measured, it is necessary for an operator to designate a measurement point while observing a photographed image and perform measurement. was there. For this reason, it took time and effort when there were many measurement points. In addition, when trying to measure with high accuracy, there is a tendency that individual differences are apt to occur and measurement is impossible.

[0013] The above situation can be avoided if the measurement is performed by attaching a mark to an object, but in such a case, as described above, additional work is required for attaching the mark. Was sometimes difficult or impossible to measure.

In the case of surface measurement, it has been necessary to manually determine a template image when performing automatic measurement by stereo matching as preprocessing, and determine an optimum search width. Further, when there is a mismatching point or the like, it is necessary to perform correction manually, which makes it difficult to achieve automation.

The present invention has been made in view of the above-mentioned problems of the prior art, and enables an efficient and high-accuracy operation from the orientation operation to the measurement, and can be carried irrespective of an object to be measured. It is intended to provide.

[0016]

According to the present invention, there is provided a rough position measuring section for roughly calculating a positional relationship of a measuring object in each image from a pair of images of the measuring object taken from different directions; In the images, one image is defined as a reference image, the other image is defined as a search image, a reference data block is set in the reference image based on the positional relationship obtained by the approximate position measurement unit, and a search area is provided in the search image. A data setting unit that sets a search data block in a search area; and a correspondence determination unit that determines a correspondence between a search data block and a reference data block that are set in the search area. At least one of setting of a reference data block in a reference image, setting of a search area in a search image, and setting of a search data block based on the positional relationship obtained by the approximate position measurement unit. It is summarized as image measuring apparatus characterized by being configured to perform.

[0017]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An image measuring apparatus according to the present invention comprises: a pair of images obtained by photographing a measuring object from different directions; Among the images, one image is defined as a reference image, the other image is defined as a search image, a reference data block is set in the reference image based on the positional relationship obtained by the approximate position measurement unit, and a search region is provided in the search image. A data setting unit that sets a search data block in the search area; and a correspondence determination unit that determines a correspondence between the search data block and the reference data block set in the search area. Based on the positional relationship determined by the approximate position measurement unit, setting of a reference data block to a reference image,
At least one of a search area setting and a search data block setting in a search image is performed.

The pair of images include a characteristic pattern image whose positional relationship is known in advance, and the approximate position measuring unit determines the position of the characteristic pattern image in the reference image and the characteristic in the search image. Based on the positional relationship between the pattern images, the configuration can be such that the positional relationship between the measurement objects in each of the pair of images is roughly obtained.

The data setting unit may be configured to set a range of a search area in the search image based on a position of the feature pattern image in the search image.

The data setting section is configured to search for a search image based on a positional relationship between a position of the feature pattern image in the search image and a position of the search area, or a position of the feature pattern image in the reference image and a reference data block position. Can be configured to set the range of the search area.

The data setting unit is configured to search for a search image based on a positional relationship between a position of a characteristic pattern image in a search image and a position of a search area, or a position of a characteristic pattern image in a reference image and a reference data block position. Can be configured to set the size of the range of the search area.

The data setting section increases the width or area of the range of the search area of the search image as the feature pattern image and the search area in the search image or the feature pattern image and the reference data block in the reference image move away from each other. Can be configured to set.

The data setting section may be configured to set a plurality of search areas of different sizes.

[0024] The data setting section may be configured to set a reference data block in the reference image based on the position of the characteristic pattern image in the reference image.

The data setting unit may be configured to set the size of the reference data block in the reference image based on the position of the characteristic pattern image in the reference image and the position of the reference data block.

The data setting section may be configured to set the height or width of the reference data block to be larger in the reference image as the position of the characteristic pattern image in the reference image and the position of the reference data block become farther apart.

[0027] The data setting unit may be configured to set a plurality of reference data blocks of different sizes based on the characteristic pattern image in the reference image.

[0028]

DESCRIPTION OF THE PREFERRED EMBODIMENTS First, an outline of an image measuring apparatus according to the present invention will be described with reference to FIG.

The image measuring device 30 includes a general position measuring unit 31,
A data setting unit 32 and a correspondence determination unit 33 are provided.

The approximate position measuring unit 31 uses a characteristic pattern image whose positional relationship is known in advance to calculate the positional relationship of the measuring object in each image from a pair of images obtained by photographing the measuring object from different directions. Is roughly obtained.

The data setting section 32 determines one of the pair of images as a reference image and the other as a search image,
The reference data block is set in the reference image and the search area is set in the search image on the basis of the positional relationship obtained in (1).

The correspondence discriminating unit 33 is configured to obtain a correspondence between a search data block set in a search area and a reference data block.

The use of the image measuring device 30 makes it possible to automate three-dimensional measurement described in detail below.
Measurement accuracy can be improved.

Hereinafter, a shape measuring apparatus using the image measuring apparatus of the present invention will be described in detail.

FIG. 1 shows a flow of measurement in the shape measuring apparatus, and FIG. 2 is a block diagram conceptually showing the configuration.

A. Stereo imaging of the object first, shed feature pattern to the measured object 0 from the feature pattern projection unit 3 by an instruction from the controller 4. Then, the image of the measurement object 0 is photographed by the left and right image photographing units 1 and 2, and the image data is transferred to the feature pattern extracting unit 5.

As shown in FIG. 3, the left and right image photographing units 1 and 2 include an optical system 11, a CCD 12, a CCD driver 13,
It comprises an operational amplifier 14, an A / D converter 15, and the like.

FIGS. 5 and 6 show examples of characteristic patterns. These characteristic patterns 20 are formed in a circular shape, but may have any shape other than the circular shape as long as the position of the mark image by the characteristic pattern projection can be obtained.

The feature pattern projecting section 3 may be any device capable of projecting the feature pattern 20, such as a slide projector or a laser pointer.

FIG. 4 is a detailed block diagram of the characteristic pattern extraction unit 5.

The image data transferred by the left and right image photographing units 1 and 2 is converted into digital data by the A / D converter 15 in FIG. 3 and transferred to the characteristic pattern projection image memory 51 of the characteristic pattern extraction unit 5. You.

Next, according to an instruction from the controller 4,
The feature pattern projection is stopped, images without a feature pattern are shot by the left and right image shooting units 1 and 2, and digital image data is transferred to the feature pattern absence image memory 52 of the feature pattern extraction unit 5.

After the image transfer to the image memory for feature pattern projection 51 and the image memory for no feature pattern 52 is completed, the image difference calculator 53 is operated by the instruction of the controller 4.
Through the two images. Then, the difference image is
It is taken into the characteristic pattern image memory 54.

As a result, the data in the feature pattern image memory 54 is only the data in which the image information of the measurement object 0 has been deleted, that is, only the information (mark image data) related to the feature pattern 20.

B. In order to perform position detection orientation processing, the position of a mark image is detected by projecting a characteristic pattern.

The characteristic pattern position detecting section 55 detects the characteristic pattern coordinate position in the characteristic pattern image memory 54.

The shape of the characteristic pattern 20 may be any shape as long as its position can be clearly understood. Here, the characteristic pattern 20 as shown in FIGS. 5 and 6 is assumed. Also, as long as there are six or more characteristic patterns as shown in FIGS. 5 and 6, any number of characteristic patterns may be used.

Since the characteristic pattern image memory 54 does not include any information other than the mark image formed by the characteristic pattern projection, erroneous detection can be eliminated. Further, since it is easy to automatically perform position detection, stable position coordinate detection can always be performed with high accuracy without individual differences.

Here, a case where the template matching method is used for detecting the approximate position of a point and the moment method is used for detecting the detailed position will be described.

As the template matching method, a residual sequential test method (SSDA method), which is a type of correlation method, will be described, but a normalized correlation method or the like may be used. Further, for the detailed position detection, a LOG filter method or the like may be used instead of the moment method.

Hereinafter, the position detecting process will be described.

(Schematic Position Detection) Register a template image.

As the template image, a simulation image similar to one mark of the projection feature pattern 20 shown in FIGS. 5 and 6 may be created, or an actual image may be selected and used.

2. A point where S> R (a, b) is searched for in the entire image (see Equation 1).

The closer the R (a, b) is to 0, the higher the similarity. S is determined in advance with an appropriate value. In this case, since the image information other than the characteristic pattern has been deleted, it can be easily determined. The template matching may use a normalized correlation method or the like, but the processing can be further speeded up by using a residual sequential test method.

[Residual sequential test method (SSDA method)] SSD
FIG. 7 shows the principle of the method A, and Equation 1 shows the equation.

The point at which the residual R (a, b) is minimum is the position of the mark to be obtained.

[0058]

(Equation 1) In the addition of the expression, if the value of R (a, b) exceeds the minimum value of the residual in the past, the addition is terminated, and the calculation processing is performed so as to move to the next (a, b), thereby speeding up the processing. Can be measured.

3. R (a, b) Minimum and adjacent mark positions are determined by the number of marks in accordance with conditions such as the distance between adjacent mark positions, and are used as position coordinates.

(Detailed Position Determination) Determination of Search Area A search area centered on a point determined by the above (approximate position detection) is set.

[0061] 2. Determination of Mark Area An area having a density value equal to or higher than a threshold value in the search area is defined as one mark (see FIG. 8).

As the threshold value, an appropriate value is determined in advance. The image is almost 0 because the difference processing has been performed on the image except for the mark.

3. Center-of-gravity position calculation The moment method is used to calculate the center-of-gravity position coordinates.

[Moment Method] As shown in FIG. 8, the following equation is applied to points (mark K) that are equal to or greater than the threshold value T.

[0065]

(Equation 2)

[0066]

(Equation 3) With the equations (2) and (3), the position of the center of gravity can be calculated up to the sub-pixel position.

4. Perform 1 to 3 for the number of points.

As described above, the position coordinates of the mark image by the characteristic pattern projection as shown in FIG. 5 or FIG. 6 can be calculated.

The detailed position detection may be performed from the beginning without performing the approximate position detection, or the position may be detected by another algorithm.

In any case, since the image is only the characteristic pattern information, the position can be calculated at high speed and with high accuracy.

[0071] c. Orientation Next, the coordinate values obtained by the characteristic pattern position detection unit 55 are sent to the posture calculation unit 6 to perform orientation calculation.

By this calculation, the positions of the left and right cameras and the like are obtained.

The parameters are obtained by the following coplanar conditional expression.

[0074]

(Equation 4) The origin of the model coordinate system is set to the left projection center, and the line connecting the right projection center is set to the X axis. The scale is based on the base line length as a unit length. The parameters obtained at this time are the rotation angle κ1 of the Z axis of the left camera, the rotation angle φ1 of the Y axis, the rotation angle κ2 of the Z axis of the right camera, the rotation angle φ2 of the Y axis, X
There are five rotation angles of the shaft rotation angle ω2. In this case, since the rotation angle ω1 of the X axis of the left camera is 0, there is no need to consider it.

Under such conditions, the coplanar conditional expression of Expression 4 is as shown in Expression 5, and each parameter can be obtained by solving this expression.

[0076]

(Equation 5) Here, the following relational expression of coordinate conversion is established between the model coordinate system XYZ and the camera coordinate system xyz.

[0077]

(Equation 6)

[0078]

(Equation 7) Using these formulas, unknown parameters are obtained by the following procedure. 1. The initial approximate value is usually set to 0. 2. Taylor expansion of coplanar conditional expression 5 around the approximate value,
The value of the differential coefficient at the time of linearization is obtained by Expressions 6 and 7, and an observation equation is established. 3. By applying the least squares method, a correction amount for the approximate value is obtained. 4. Correct the approximation. 5. Using the corrected approximate value, the above-mentioned 2. ~ 5. Is repeated until convergence.

If some of the marks are not properly detected due to the shape of the object or the like, the convergence may not be achieved. In such a case, the above steps 1 to 5 are performed by deleting one point at a time, and the converged or best parameter is used.

[0080] d. Stereo Model Formation Next, the image is transformed into a stereo model coordinate system that can be viewed stereoscopically by using the parameters obtained by orientation, and a stereo model is formed.

The conversion formula to the model coordinates is as follows.

[0082]

(Equation 8)

[0083]

(Equation 9)

[0084]

(Equation 10) Thus, three-dimensional measurement is possible.

[0085] e. Three-dimensional measurement (1) Point measurement By using a feature pattern 20 having six or more points as shown in FIG. 6, a plurality of marks are subjected to the same processing as in position detection (b), so that their three-dimensional coordinates are automatically increased. Required for accuracy. (2) Surface Measurement By using the projection feature pattern 20 as shown in FIG. 6, it is possible to set an initial value for stereo matching.

How to determine the initial value in the case of surface measurement will be described.

FIGS. 9 and 10 are explanatory diagrams showing a part of the projection feature pattern as shown in FIG. FIG. 9 is a reference image as a reference, and FIG. 10 is a search image to be searched. These reference / search images may be left or right images. For example, the left image is determined as a reference image, and the right image is determined as a search image.

The projected feature patterns are the reference images S′1, S ′
On the search image, S1, S2,
S5 and S6 correspond.

The corresponding point search is performed by performing stereo matching on a search area in the search image using the reference data block in the reference image as a template image. Image correlation processing or the like is used for stereo matching.

[Determination of Search Area] A method of determining a search area based on a projected feature pattern will be described.

Projection feature patterns S1 and S2 (horizontal direction)
Is a width A, S1 to S5 (vertical direction) is a width B, and S1, S
2. A portion surrounded by S5 and S6 is defined as a search region R1.
Similarly, a portion surrounded by S2, S3, S6, S7 is determined as a search region R2,..., And a portion surrounded by the respective projected feature patterns is determined as a search region.

As shown in FIG. 11, the mark images based on the actual projected feature patterns are not always on the same line depending on the shape of the object and the direction of the CCD. Therefore, from the four points of each projection feature pattern, the maximum A,
A rectangle that can obtain the width of B is created as a search area. For example, in FIG. 11, the maximum widths that can be taken in S1, S2, S5, and S6 are the horizontal direction A1 and the vertical direction B1.
A rectangle created with the maximum widths A2 and B2 that can be taken by S2, S3, S6, and S7 is defined as a search region R2. In this way, although there is a somewhat overlapping area, R1, R
The boundary between the two regions can also be reliably searched.

This is to determine the range of the search area in the search image based on the position of the mark image in the search image that has already been obtained.

FIG. 12 shows a modification of the search area determination.

For example, each projection feature pattern is
Since it is already required as a corresponding point on the left and right images,
It is efficient if the search start position and the end position of the search area are set as areas near the projection point. That is, in the example of FIG. 11, as the search proceeds in the vertical direction from S1 to S5 by the search width A1, the useless search area (the area where no corresponding point exists) becomes closer to the vicinity of the line of S5. Out.

Therefore, the same horizontal point as S1 is set as the start point of the search area up to D1 which is 1/2 of the interval between S1 and S5 (vertical direction) in FIG. 12, and D1 and S5 are set between D1 and S5.
Take the horizontal search start point on the line connecting.

If such processing is set as a search area for each projection point and stereo matching is performed while slightly overlapping the search areas, a somewhat efficient search can be performed.

FIG. 13 and FIG. 14 show that the search area is further improved by setting search data blocks in the search area.

For example, marks S1 to S2 in a search image
In the section A1, the positions of the reference data blocks T1, T2, T3,... On the reference image shown in FIG. 9 in the search area corresponding to the respective search data blocks are obtained as shown in FIG. , U1, U2,
U3,... Are set in several blocks. By doing so, the time in the search area search can be shortened and the search can be performed efficiently. In this case, the range of the search data block can be determined by the number of reference data blocks. For example, if n reference data blocks are set between A's of the reference images S'1 to S'2, A1 / n or A1 / (n-1) may be used. (N-1) is the case where the search is performed by slightly overlapping the search data blocks.

In this method, the range of the search area is set in the search image based on the position of the mark image and the position of the search area in the search image that have already been obtained. In other words, it can be said that the range of the search area is set in the search image based on the positional relationship between the position of the mark image in the reference image and the position of the reference data block.

FIG. 14 shows a modification of the system shown in FIG. This makes the size of the search data block variable. That is, since it is known that the corresponding point with respect to the reference data block is near in the neighboring area of the marks S1 and S2, the size of the search data block is small, and the position of the corresponding point becomes uncertain as the distance increases, so that the size is increased. It is. Therefore, the reference data block has the largest search data block size at the position of A '/ 2.

The size of these search data blocks is S
It is determined from the search area A1 of 1 to S2.

For example, if n is the number of reference data blocks between A ′, search data block size in the section of S1 to S1 + A1 / 2: (1 + t × i / n) × A1 / n, S1 + A1 / 2− Search data block size in the section of S2: (1 + t × (ni) / n) × A1 / n, where i is the position of the search data block corresponding to each reference data block, ie, i = 1 to n. I do.
Further, t is a constant of the magnification and is set to a predetermined value. For example, S
If it is desired to double the search data block size at the position (U1) of S1 at the position of 1 + A1 / 2, select 2.

The search data block size may be determined based on the marks S′1, S′2 in the reference image and the reference data blocks T1, T2,. In this case, A1 is set to A ', and the term is multiplied by a term in consideration of the magnification A1 / A'.

As described above, by setting the search area and the search data block, it is possible to perform the corresponding point search efficiently, that is, at high speed and with high reliability.

In this method, if the viewpoint is changed based on the positional relationship between the position of the mark image in the search image and the position of the search area, the positional relationship between the position of the mark image in the reference image and the position of the reference data block is changed. Based on the search image, the size of the range of the search area is set.

[Determination of Reference Data Block] Reference data blocks serving as references are S'1, S'2,
S'5, S'6, or S1, S2, S on the search image
5. Determined from S6. For example, the reference image (FIG. 9) S'1
If S'2 is A 'and S'1 to S'5 are B', the reference data block width in the horizontal direction can be determined as A '/ n and in the vertical direction as B' / m.

Alternatively, the size may be proportional to the ratio of the left and right images. For example, the horizontal direction is set as A '/ A * n, and the vertical direction is set as B' / B * m.

Although n and m can be determined as appropriate in accordance with the relationship between the size of the object to be measured and the number of pixels, they may be constants as appropriate according to the values of A 'and B'.

FIG. 15 shows a method of performing stereo matching in the same manner as the above-described processing while obtaining correlation products with three types of reference data block sizes instead of one type. Also in this case, these three sizes can be determined based on the information of A 'and B'.

FIG. 16 shows a further modification of the reference data block determination according to the present invention. In the regions near S′1 and S′2, the reference data block may be small because the correspondence is relatively high on the reference / search image, but the corresponding position becomes uncertain as the distance increases, so that the reference data block becomes uncertain. Change size dynamically. For example, the reference data block is enlarged at the T1, T2, and T3 positions as shown in FIG. Then, the reference data block position is set to the maximum size at the point A '/ 2. A '/
From 2 to S′2, the reference data block size is gradually reduced in reverse.

By doing so, the reliability of stereo matching can be improved.

These reference data block sizes are S ′
The width is determined from the horizontal width A 'of 1 to S'2 and the vertical width B' of S'1 to S'5.

For example, if n and m are the number of reference data blocks between A 'and B': Reference data block size in the section of S'1 to S'1 + A '/ 2: Horizontal direction: (1 + t × i / N) × A ′ / n, vertical direction: (1 + t × l / m) × B ′ / m ・ Standard data block size in the section from S′1 + A ′ / 2 to S′2 Horizontal direction: (1 + t × (n −i) / n) × A ′ / n, vertical direction: (1 + t × (mi) / m) × B ′ / m, where i and l are respective reference data block positions, that is, i = 1 to n , 1 = 1 to m.

Here, t is a magnification constant, and is set to a desired value. For example, if the position of S'1 + A '/ 2 is to be double the reference data block size of the position (T1) of S'1, "2" is selected.

In this way, the reference data block can be made variable.

Further, if the three types of reference data blocks as shown in FIG. 15 are made variable as in the above-described example and the correlation products are obtained and stereo matching is performed, a more reliable corresponding point search can be performed. Becomes

The determination of these reference data blocks may be similarly performed based on the marks S1 and S2 in the search image. In this case, A1 is A1 and B 'is B1. Further, the reference image and the search image are multiplied by a term in consideration of the magnification A' / A1.

As described above, each initial value is automatically obtained.

Next, the procedure of actual measurement (stereo matching) will be described.

As an example, the projected feature patterns S1 to S in FIG.
The case of No. 8 will be described with reference to FIGS.

In the vertical direction, since the orientation processing has been completed and the vertical parallax has been removed (aligned), the reference image,
It is only necessary to search on the same line of the search image. When setting the search data block, the search data block is set appropriately as described above in each search area.

1. For the search width A1 of the horizontal line L1 in the search area R1 from S1 to S2 of the search image, corresponding point search is sequentially performed on the reference data blocks T1, T2, T3,.

These initial values are determined by the method described above.

When A1 on line L1 of search area R1 is completed, For the search area R2 of S2 to S3, the corresponding point search on A2 of the line L1 is sequentially repeated from the reference data block position determined in this area.

3. Next, a corresponding point is searched for the line A3 of the line L1 in the search area R3 of S3 to S4. For the search area R3, the search is sequentially performed with the template size and position determined in this area.

When the search for the corresponding point on the line L1 is completed, the process moves to the next line L2, and the corresponding point search is repeated from the search width A1 of the search area R1 in the same manner as in steps 1 to 3.

This is repeated for the required number of lines.

As described above, the initial values for stereo matching, that is, the search area, the search data block, and the reference data block are automatically determined from the position of the projection feature pattern, and automatic measurement can be performed. .

Further, since the search area and the reference data block of the reference / search image can be limited to within A ′ and B ′, processing can be performed at a much higher speed than ordinary stereo matching. That is, usually, stereo matching for one horizontal line (line in FIG. 6) is performed at each search position, but the processing time is reduced to a fraction of the number of projection points.

Further, since the corresponding area on the reference / search image is previously determined and determined, mismatching can be greatly reduced. That is, the reliability of stereo matching can be significantly improved.

In addition, since the reference data block, the search area, and the search data block can be appropriately determined based on the position information of the projection point, the reliability of stereo matching is further improved.

Further, since the reference data block and the search data block can be dynamically changed according to the search position, the reliability is further improved.

Since the stereo matching can be performed on an image having no projection feature pattern, erroneous detection due to the feature pattern does not occur. An image without a feature pattern has a value as an image database, and an original image can be stored simultaneously with measurement.

Further, according to the above-described embodiment, since an image on which the characteristic pattern is projected and an image on which the characteristic pattern is not projected are used, it is not necessary to attach a measurement mark to the object, and the mark is not attached. Measurement can be performed on an object.

As described above, since an image having only the feature pattern information can be created by subtracting the image having the feature pattern from the image having no feature pattern, the process from the feature pattern position detection to the orientation to the stereo model formation to the three-dimensional measurement can be performed manually. This makes it possible to automatically and accurately perform the operation regardless of the above. That is,
Conventionally, complicated and complicated manual orientation work and three-dimensional measurement work can be eliminated, and all operations can be performed automatically to improve reliability.

Further, it is not necessary to firmly fix the cameras on the stereo gantry, and high-precision three-dimensional measurement can be performed by simply installing two cameras roughly at the site and photographing. There is an excellent effect that measurement can be easily performed irrespective of the object or the object.

Further, by increasing the number of feature pattern points, it becomes possible to set the initial value of stereo matching, that is, to automatically determine a search width and a template image and perform stereo matching. There is an outstanding effect that the surface shape of the object can be automatically measured while significantly improving the reliability.

[0139]

According to the present invention, the positional relationship of a measurement object in each image is roughly determined from a pair of images of the measurement object taken from different directions, and one of the pair of images is used as a reference. An image, the other image is determined as a search image, a reference data block is set in the reference image based on the positional relationship obtained by the approximate position measurement unit, a search area is provided in the search image, and the search data block is located in the search area. In order to determine the correspondence between the search data block set in the search area and the reference data block, the reference data block is set in the reference image based on the approximate positional relationship, and the search area is set in the search image. By setting at least one of the setting of the search data block and the setting of the search data block, the correspondence can be accurately obtained.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a measurement flow in a shape measuring apparatus according to the present invention.

FIG. 2 is a block diagram conceptually showing a shape measuring apparatus of the present invention.

FIG. 3 is a conceptual diagram showing left and right image photographing units of the shape measuring device.

FIG. 4 is a block diagram showing a pattern extracting unit of the shape measuring device.

FIG. 5 is a plan view showing an example of a projection pattern.

FIG. 6 is a plan view showing another example of the projection pattern.

FIG. 7 is an explanatory diagram showing template matching by the SSDA method.

FIG. 8 is an explanatory diagram showing a moment method.

FIG. 9 is an explanatory diagram showing a reference image.

FIG. 10 is an explanatory diagram showing a search image.

FIG. 11 is an explanatory diagram showing a search image.

FIG. 12 is an explanatory diagram showing a search image.

FIG. 13 is an explanatory diagram showing a search image.

FIG. 14 is an explanatory diagram showing a search image.

FIG. 15 is an explanatory diagram showing a reference image.

FIG. 16 is an explanatory diagram showing a reference image.

FIG. 17 is a block diagram showing an outline of an image measuring device according to the present invention.

[Explanation of symbols]

 0 Measurement object 1, 2 Left and right image photographing unit 3 Pattern projecting unit 4 Controller 5 Pattern extracting unit 6 Attitude calculating unit 7 Stereo model forming unit 8 Shape measuring unit 9 Template matching image 10 Shape measuring device 11 Optical system 12 CCD 13 CCD driver Reference Signs List 14 Operational amplifier 15 A / D converter 20 Projection pattern 30 Image measuring device 31 Rough position measuring unit 32 Data setting unit 33 Correspondence determining unit 51 Image memory for pattern projection 52 Image memory for no pattern 53 Image difference calculation unit (unit) 54 pattern Image memory 55 Pattern position detector

Claims (11)

[Claims] 1. A schematic position measuring unit for roughly calculating a positional relationship of a measurement target in each image from a pair of images obtained by photographing the measurement target from different directions, and one image in the pair of images is used as a reference image. The other image is defined as a search image, a reference data block is set in the reference image based on the positional relationship obtained by the rough position measurement unit, a search area is provided in the search image, and a search data block is set in the search area. A data setting unit, and a correspondence determination unit that determines a correspondence relationship between a search data block set in the search area and a reference data block, wherein the data setting unit determines a positional relationship obtained by the approximate position measurement unit. Based on the setting of the reference data block in the reference image, setting of the search area in the search image and setting of the search data block Characteristic image measuring device. 2. The image measurement apparatus according to claim 1, wherein the pair of images include a characteristic pattern image whose positional relationship is known in advance, and the approximate position measurement unit includes On the basis of the position of the characteristic pattern image and the positional relationship of the characteristic pattern image in the search image, the positional relationship of the measurement target in each of the pair of images is roughly obtained. Image measuring device. 3. The image measurement device according to claim 2, wherein the data setting unit is configured to set a range of a search area in the search image based on a position of a feature pattern image in the search image. An image measuring device characterized by the above-mentioned. 4. The image measurement device according to claim 2, wherein the data setting unit is configured to determine a position of a characteristic pattern image and a position of a search area in a search image or a position and a reference data of a characteristic pattern image in a reference image. An image measuring device configured to set a range of a search region in a search image based on a positional relationship with a block position. 5. The image measurement apparatus according to claim 2, wherein the data setting unit is configured to determine a position of the characteristic pattern image and a position of the search area in the search image, or a position and a reference data of the characteristic pattern image in the reference image. An image measuring apparatus, wherein a size of a range of a search area is set in a search image based on a positional relationship with a block position. 6. The image measurement device according to claim 5, wherein the data setting unit searches as the feature pattern image and the search area in the search image or the feature pattern image and the reference data block in the reference image move away. An image measuring apparatus, wherein the width or area of a range of an image search area is set to be large. 7. The image measurement apparatus according to claim 3, wherein the data setting unit is configured to set a plurality of search areas having different sizes. measuring device. 8. The image measurement device according to claim 2, wherein the data setting unit is configured to set a reference data block in the reference image based on a position of the feature pattern image in the reference image. Image measuring device. 9. The image measurement device according to claim 2, wherein the data setting unit determines the size of the reference data block in the reference image based on the position of the characteristic pattern image in the reference image and the reference data block position. An image measurement device characterized by being configured to set. 10. The image measurement device according to claim 2, wherein the data setting unit is configured such that the height of the reference data block or the height of the reference data block is increased as the position of the characteristic pattern image in the reference image and the position of the reference data block become farther apart. An image measuring apparatus characterized in that the width is set to be large. 11. The image measurement device according to claim 8, wherein the data setting unit sets a plurality of reference data blocks having different sizes based on a characteristic pattern image in the reference image. An image measuring device characterized by having such a configuration.