JP2004085371A - Non-contact shape measuring arrangement - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a non-contact shape measuring arrangement which can solve a problem of data dropout caused by insufficient light intensity and record detailed color information. <P>SOLUTION: Three dimensional data of a measurement object 10 is calculated on the basis of moving positions of the measurement object 10 and a planar coordinate system calculated for each line image. Images of the object 10 are picked up by a camera 20 from a main camera position 20A and an auxiliary camera position 20B to obtain central projection images. A pseudo central projection image is constituted by using the three dimensional data of the object calculated as above and imaging data such as the position and attitude of the camera 20. The pseudo central projection image is compared with the central projection image acquired at the main camera position 20A to determine a site on which the central projection image lacks dada. The site is interpolated by using a three-dimensional image made of the central projection images of the site picked up at the position 20A and the position 20B. <P>COPYRIGHT: (C)2004,JPO

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a measuring device using a high-resolution image, and more particularly to an interpolation method (data processing method) for a data missing portion of a non-contact type measuring device, acquisition of high-resolution image information, and improvement of a measuring device for realizing the method. .
[0002]
[Prior art]
The object to be measured (measurement object) is irradiated with slit light from a slit light source, and the projected slit image of the object to be measured is imaged at a predetermined angle or at an equal pitch by an imaging means, and the three-dimensional image of the object to be measured is obtained. Various techniques for measuring the shape have been proposed (see JP-A-3-138507, JP-A-5-203423, and JP-A-9-102913).
[0003]
As this type of non-contact type three-dimensional shape measuring apparatus, an apparatus using a line laser, a uniaxial moving table, and a CCD camera has already been proposed (see Japanese Patent Application Laid-Open No. 11-183125). As shown in FIG. 6, in this three-dimensional shape measuring apparatus, the light surface (light surface) is constituted by the XY plane by the slit light beams b1 and b2 emitted from the left and right laser light sources a1 and a2. Then, the measurement object d placed on the uniaxial moving table c is passed through the light surface in the Z direction, and the image of the line image of the portion of the measurement object d that hits the slit light is displayed on the front surface of the measurement object d ( The image is captured by a CCD camera e provided on the front side in the Z direction). Note that the captured image data is an analog signal, which is converted into a digital signal via the analog / digital conversion converter f and input to the computer g, converted into an analog signal by the digital / analog conversion converter h, and processed on the television monitor. Is displayed.
An image obtained when the measurement object d passes through the light surface acquires cross-section information (hereinafter, referred to as a “cross-section image”) of the measurement object d as a line image representing a partial outer shape of the measurement object d. However, the three-dimensional shape is measured by calculating the plane coordinates of the cross-sectional portion from the cross-sectional image and adding the movement amount of the uniaxial moving table c as depth information.
In addition, by superimposing each section information and expressing a bright color with priority, creation of an orthographic projection image (hereinafter, referred to as “appearance image”) is realized.
[0004]
[Problems to be solved by the invention]
However, in the invention described in the above publication, only the portion that has passed through the surface of the light constituted by the slit light is converted into data. Therefore, in the state shown in FIG. Also, there is a disadvantage that the slit light does not reach the leading end portion j2 and data is lost due to insufficient light quantity. FIG. 8 shows an example of data loss due to insufficient light quantity. In order to remedy this drawback, a color adjustment (luminance value correction) method by image processing of the portion is proposed by the technique described in the above-mentioned publication. However, the above technique cannot solve the drawbacks such as the fact that data itself is not created due to insufficient light quantity.
[0005]
SUMMARY OF THE INVENTION The present invention has been made to solve the above-mentioned conventional problems, and eliminates the problem of data loss due to insufficient light quantity, and can record detailed color information. The purpose is to provide.
[0006]
[Means for Solving the Problems]
The present invention captures an image of a measurement target irradiated with slit light, and captures, as a line image, line light that appears in a line on the surface of the measurement target by the slit light, and the measurement target object Calculates the plane coordinates of each line image on the measurement surface of the object to be measured based on each line image obtained by the first imaging unit each time the object moves relative to the first imaging unit at a predetermined interval. A three-dimensional data calculating unit that calculates three-dimensional data of the surface of the measurement target based on each movement position of the measurement target detected by the position detection unit and the plane coordinates calculated for each line image; A second imaging means for photographing the object to be measured from its front position to obtain a central projection image, and photographing the object to be measured at a lateral position at a certain interval from the front of the object to be measured A third imaging means for obtaining a center projection image by the second imaging means based on the calculated three-dimensional data of the measurement object and shooting information such as the position and orientation of the first imaging means. Means for creating a pseudo center projected image similar to the acquired center projected image, and comparing the pseudo center projected image with the center projected image to determine a portion where the data of the pseudo center projected image is missing and should be interpolated Based on the center projection image obtained by the second imaging unit, the center projection image captured by the third imaging unit, and shooting information (for example, imaging unit information) such as the position and orientation of each imaging unit. A non-contact type shape measuring apparatus comprising: a calculating means for calculating a three-dimensional shape of a part to be interpolated and interpolating the part.
According to the present invention, a pseudo center projection image is created based on three-dimensional data of a measurement target and shooting information such as the position and orientation of the first imaging unit. In this pseudo center projected image, as shown in FIG. 7, since the data is missing in the convex portion or the concave portion where the slit light does not hit, in order to identify the missing portion, the pseudo center projected image and the central projected image are used. To determine the data missing portion. Then, based on the center projection image acquired by the second imaging unit, the center projection image photographed by the third imaging unit, and imaging information such as the position and orientation of each imaging unit, the three-dimensional shape of the part to be interpolated is determined. By calculating and interpolating the relevant part, it is possible to obtain three-dimensional data (three-dimensional position data) in perfect form, that is, three-dimensional shape data.
In the present invention, the three-dimensional shape data that projects the pseudo center projection image is used by using the color information of the center projection image acquired by the second imaging means as the color information of the pseudo center projection image for which interpolation has been completed. It is preferable to provide a means for giving the central projection color information to the image data. In this way, it is possible to give the center projection color information to the three-dimensional shape data constituting the pseudo center projection image, and as a result, to create three-dimensional shape data having detailed color information and an appearance image Can be.
[0007]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, an embodiment of a non-contact type three-dimensional shape measuring apparatus according to the present invention will be described with reference to the drawings.
FIG. 1 is a configuration explanatory view showing one embodiment of a non-contact type three-dimensional shape measuring apparatus according to the present invention. As shown in FIG. 1, the measurement target is placed on the measurement table 12. In FIG. 1, a measurement object 10 is an archeological object excavated such as a Jomon pottery, but the measurement object 10 is not limited to an archeological excavation item, but may be another article.
[0008]
The measurement table 12 on which the measurement object 10 is placed is provided so as to be movable in one axis direction (Z direction) by a driving mechanism (not shown). This drive mechanism receives a drive signal from a personal computer (hereinafter abbreviated as “PC”) 24 described later and moves the measuring table 12 back and forth in the Z direction. Therefore, the measurement object 10 on the measurement table 12 moves in the Z direction along with the movement of the measurement table 12 by the drive signal from the personal computer 24.
[0009]
Light sources 14 and 16 are arranged symmetrically on both sides of the measuring table 12. The light sources 14 and 16 are planar in a plane represented by (X, Y) coordinates that intersect perpendicularly with the Z direction, which is the moving direction of the measurement table 12 (hereinafter, this plane is referred to as a measurement plane). The illumination light that spreads out. The light sources 14 and 16 can emit illumination light such as laser light from a point light source through a slit, for example. Depending on the shape of the measurement target, it is also possible to install a light source at a position other than a symmetric position such as the upper direction.
As a result, the illumination light is applied linearly to the surface of the measurement object 10 from both sides in the measurement plane (hereinafter, this illumination light is referred to as slit light). The portion illuminated by the slit light shows the contour shape (line image) of the cross section when the measurement object 10 is cut at the measurement surface.
[0010]
At the time of acquiring the line image, the light other than the slit light emitted from the light sources 14 and 16 is shielded around the measuring device, so that the light other than the surface of the measurement target 10 to which the slit light is irradiated is irradiated. Is not irradiated with light. Marks 18A, 18B, 18C and 18D indicating reference points for determining plane coordinates in the X and Y directions in FIG. These marks 18A, 18B, 18C, and 18D have a clear positional relationship with each other, and serve as a reference when detecting the plane coordinates of the contour shape of the measurement object 10 on the measurement surface, as described later. Also, the screen distortion is corrected by detecting the positions of the marks 18A, 18B, 18C, 18D.
[0011]
On the other hand, in front of the measurement table 12 in the Z direction, a high-resolution digital camera is directly opposed to the measurement object 10, and the camera 20 installed in front of the measurement object as in the conventional FIG. (Hereinafter, referred to as a main camera position 20A), and slidably (movable) in a horizontal direction at regular intervals before continuous shooting of a slit image (hereinafter, referred to as an auxiliary camera position 20B). Before the continuous photographing of the line image by the camera 20, when the camera 20 is at the auxiliary camera position in the lateral direction (corresponding to a third imaging means), the measurement target placed at the photographing start position of the measuring table 12 A test photographing is performed to photograph the images 10 and obtain marks 18A, 18B, 18C, and 18D, which serve as references for the length and the coordinate system, to acquire a projection image. Further, the camera 20 is returned to the main position, and the measurement object 10 placed at the photographing start position of the measuring table 14 is photographed to obtain a central projection image, and a mark 18A serving as a reference for the length and the coordinate system, Test photography for photographing 18B, 18C, and 18D is performed.
When the center projection image, the projection image, and the marks 18A, 18B, 18C, and 18D are photographed (at the time of test photographing), the camera 20 performs lighting so that the measurement object 10 does not generate shadows or highlights as a whole. It is preferable to narrow down the focus so that the entirety is in focus.
[0012]
Further, during continuous shooting of line images, the camera 20 adjusts the focus so that the position of the measurement surface is in focus, and moves the measurement target 10 and the marks 18A, 18B, 18C, and 18D from a direction perpendicular to the measurement surface. Shoot. Thus, the slit light reflected by the measurement target 10 is received by the pixel corresponding to the reflection position in the CCD camera 20. Therefore, the contour shape of the cross section of the measurement object 10 on the measurement surface can be detected from the image obtained by the camera 20.
In the embodiment, one digital camera uses a single camera to capture a projection image at a front position (main camera position 20A) and a lateral position (auxiliary camera position 20B) and a line image at a front position. However, two digital cameras can be provided to shoot at the front position and the auxiliary position, respectively. Further, a camera for taking a line image may be provided separately from the camera for acquiring the projection image. Further, in the prior art shown in FIG. 6, a CCD camera is mentioned as an image input means, but various means such as a digital camera and a video camera can be used as long as the image pickup means can obtain a digital image. Is not limited.
[0013]
The image captured by the camera 20 is input to the personal computer 24 as a digital image signal (hereinafter, referred to as image data). The personal computer 24 outputs support for moving the position of the camera 20 at regular intervals between the main camera position 20A and the auxiliary camera position 20B at the time of test photographing of the measurement object 10, and outputs a shutter control signal to the camera 20 for each position. To shoot a projected image. In addition, when measuring the external shape of the measurement object 10, the driving signal is output to the driving device of the measurement table 12 to move the measurement table 12 as described above, and the measurement table 12 is moved from the reference position. The amount is detected, and, for example, the measurement object 10 is moved from the rear to the front by a predetermined interval within a photographing range (measurement range). Further, digital image data sent from the camera 20 is input each time the measuring table 12 is moved at a predetermined interval. The moving interval of the measuring table 12 is, for example, the width of the slit light emitted from the light sources 14 and 16.
[0014]
As a result, image data indicating the contour shape of the measurement object 10 is obtained at predetermined intervals in the Z direction. In the following, images of the measurement object 10 obtained at predetermined intervals in the Z direction are referred to as cross-sectional images. The personal computer 24 obtains the X and Y plane coordinates of the surface of the measurement object 10 in each cross section based on the image data of these cross section images, and the position (reference position) of the measurement table 12 when the image data is obtained. ), The coordinates in the Z direction are obtained based on the amount of movement from the. As a result, the three-dimensional coordinates of the surface of the measurement target 10 in each cross section can be obtained, and the three-dimensional coordinates of the surface of the measurement target 10 in each cross section obtained in this manner are collectively collected. It is possible to detect the three-dimensional coordinates of each point in the (photographing range).
[0015]
The personal computer 24 performs data processing such as light amount adjustment on the image data of the cross-sectional image at the same time as detecting the three-dimensional coordinates, and synthesizes the image data of each cross-sectional image, thereby forming an image of the appearance image of the measurement object 10. Generate data. The image data thus generated is output to a TV monitor or the like via a D / A converter, and an appearance image can be displayed on the screen of the TV monitor.
[0016]
Next, a shape measuring procedure by the non-contact type three-dimensional shape measuring device will be described with reference to a flowchart of FIG. First, test photography of the measurement object 10 is performed (step S10). In this test photographing, the measuring object 10 is placed at the photographing start position during continuous photographing, and the camera 20 installed in front of the measuring object 10 is used as a main position to image the measuring object (main camera). Is slid (moved) in the horizontal direction by a certain distance (auxiliary camera) to capture an image. In each case, the marks 18A to 18D are photographed.
[0017]
Then, the measuring table 12 is moved from the photographing start point of the measuring object 12 (step S14), and a cross-sectional image of the measuring object is photographed by the camera 20 at every predetermined interval from the photographing start position (step S16). The personal computer 24 inputs image data of a cross-sectional image photographed by the camera 20 for each movement at a predetermined interval, and first detects reference point marks 18A, 18B, 18C, 18D from the input image data. Then, the distortion of the photographing optical system of the CCD camera 20 is detected from the positional relationship between the marks 18A, 18B, 18C, 18D, and the image data is processed to correct the image distortion (step S18). This correction improves the measurement accuracy. In addition, if the captured image is captured darker than the reference brightness and the image is partially missing, the brightness (brightness value) of only this portion can be raised to the reference value. .
[0018]
After performing the above-described image data correction processing, the plane coordinates of the contour shape of the measurement target 10 in the cross section are calculated from the image data of the cross section image (step S20). That is, a point of a pixel brighter than a predetermined luminance is detected in the image data of the cross-sectional image, and coordinates on the measurement surface corresponding to the pixel point are calculated. Further, the movement amount is detected from the photographing start point of the measuring table 12, and the coordinates in the Z direction (depth direction) of the cross-sectional image in the coordinate system fixed to the measurement target 10 are added to the plane coordinates based on the movement amount. (Step S22). Thereby, based on the coordinates on the measurement surface and the coordinates in the Z direction, it is possible to obtain three-dimensional coordinates of the contour shape of the measurement target 10 in the cross section where the image is captured. As described above, the brightness can be increased as a countermeasure against the shot image being shot dark.
[0019]
Next, the photographed cross-sectional images are combined (step S24). That is, the brightness (brightness value) is compared between each pixel of the photographed cross-sectional image and the basic image whose initial value is solid black, and the brightness value of the bright pixel is arranged in the reference image. Thereby, the appearance image (front image) of the measurement object 10 can be synthesized from each tomographic image of the measurement object 10. Note that, in synthesizing the cross-sectional images, a basic image whose initial value is solid black is set, and a captured image is obtained. This processing corresponds to the cross-sectional image acquisition processing in step S16 in FIG. Then, after obtaining the captured image, the brightness (luminance value) of each pixel of the basic image and the captured cross-sectional image is compared, and the luminance value of a pixel brighter than the basic image is arranged in the basic image. Thereby, the appearance image (front image) of the measurement object 10 can be synthesized from each cross-sectional image.
[0020]
After the above processing is completed, it is determined whether or not the measurement table 12 is at the photographing end position (step S25). In the case of NO, the measurement table 12 is moved forward by the width of the slit light (step S26), and the processes of steps S16 to S25 are repeatedly executed. On the other hand, in the case of YES, the photographing (measurement) is terminated (step S27), and the projection images of the main camera and the auxiliary camera of the measurement object 10 created in step S10 and calculated in steps S22 and S24. The contour shape of each section, that is, the three-dimensional coordinates of the surface of the measurement object 10 (that is, the three-dimensional coordinates of the surface of the measurement object 10, three-dimensional shape data) is temporarily recorded on an external recording medium of the personal computer 24 (step S28). ).
[0021]
Next, an image (pseudo-center projection image) having the same content as that taken at the main camera position is created based on the synthesized three-dimensional appearance image (orthographic projection image) and shooting information such as the position and orientation information of the camera. (Step S30). Then, before the start of the continuous photographing, the central projection image photographed by the main camera in the test photographing in step S10 is compared with the data missing portion to be interpolated (step S32). As shown in the missing part determination process in FIG. 4, the pseudo center projected image has a portion where data is missing due to insufficient light quantity, and the pseudo center projected image is compared with the center projected image taken by the test to perform image processing. The data missing part can be easily determined by calculating the difference between the luminance values.
Next, for a part to be interpolated, a three-dimensional shape (three-dimensional position information) is calculated and recorded using two images taken at the main camera position and the auxiliary camera position and the position and orientation information of each camera. This operation is performed for all the parts to be interpolated. Based on the three-dimensional shape calculated in this way, a three-dimensional shape obtained by synthesizing the cross-sectional shape is interpolated (step S34). In this case, as described later, the color information of the center projected coordinates is given as the color information of the pseudo center projected coordinates. The appearance recording image after interpolation and the three-dimensional shape are recorded (step S36).
[0022]
The interpolation work of steps S28 to S36 will be described in detail.
First, the appearance recording image and the three-dimensional shape are temporarily stored (step S40). Next, a pseudo center projected image is created using the three-dimensional shape information and camera information (information such as the posture and position of the camera) (step S42). The test photographed image (center projected image) at the main camera position is compared with the pseudo center projected image (step S44). By comparing the color information between the center projected image and the pseudo center projected image, a portion having a difference in color (data missing portion) is determined (step S46).
As a result of the determination, if there is a data missing portion (step S48: YES), the missing portion is specified (step S50). Next, a three-dimensional shape is calculated from two test images obtained by shooting only the data missing part in the test shooting (step S52). An embedding process (interpolation process) is performed on all of the specified parts based on the calculated three-dimensional shape (step S54). Various techniques can be used for the interpolation process in this case, and details thereof will be omitted. Examples of known interpolation processes include, for example, “Authors: Tetsuharu Anai, Hirofumi Chinzu, Article Name:“ Modeling and Virtual Application to Museums ”, published literature: Photographic Survey and Remote Sensing, Vol. 37, No. 5, pp. 47-53, published by the Photographic Society of Japan, published in 1998”.
Next, a pseudo center projected image is re-created from the interpolated three-dimensional shape (step S56). Then, the color information of the center projected image is transferred to the pseudo center projected image (step S58). Next, an appearance image is created from the pseudo center projected image to which the color information has been transcribed (step S60). The created appearance image and three-dimensional shape are recorded (step 62).
[0023]
The above-mentioned interpolation work is performed for all the parts to be complemented. Further, by copying the color information of the center projected image as the color information of the pseudo center projected image after the completion of the interpolation work, it becomes possible to give color information to the three-dimensional shape data constituting the pseudo center projected image, and as a result, detailed An appearance image can be created from three-dimensional shape data having color information. FIG. 5 shows a pseudo-center projected image of the measurement target object 10 in FIG. 4 and an appearance image (orthographic projected image) obtained by processing the center projected image.
[0024]
【The invention's effect】
As described above, according to the present invention, in non-contact type shape measurement, there is no data loss due to insufficient light quantity, and detailed color information can be recorded.
That is, according to the non-contact type shape measuring device of the present invention, the missing portion is searched by comparing the pseudo center projected image created based on the three-dimensional shape with the center projected image, and the data of the portion is interpolated. In addition, partial omission of data due to insufficient light quantity can be eliminated. Thereby, the data loss of the conventional device can be eliminated. Further, there is an effect that the operation time is reduced as compared with the method of calculating all three-dimensional images from a plurality of images used in the interpolation operation.
In addition, by using the center projection image, it is possible to acquire more detailed color information as compared with a conventional measurement result using an image illuminated by a laser.
[Brief description of the drawings]
FIG. 1 is a configuration explanatory view showing one embodiment of a non-contact type three-dimensional shape measuring apparatus according to the present invention.
FIG. 2 is a flowchart illustrating a shape measurement procedure performed by the non-contact type three-dimensional shape measurement device in FIG. 1;
FIG. 3 is a flowchart illustrating a procedure of an interpolation operation in the shape measurement procedure of FIG. 2;
FIGS. 4A and 4B are explanatory diagrams of an image for data missing determination by the non-contact type three-dimensional shape measuring apparatus according to the embodiment, wherein FIG. 4A is an explanatory diagram of an example of a pseudo center projected image, and FIG. FIG.
FIG. 5 is an explanatory diagram of an example of appearance image data created by the non-contact type three-dimensional shape measuring apparatus of the embodiment.
FIG. 6 is an explanatory diagram of an example of a conventional three-dimensional shape measuring apparatus.
FIG. 7 is an explanatory diagram of a situation of occurrence of data loss due to insufficient light quantity.
FIG. 8 is an explanatory diagram of an example of an appearance image including missing data.
[Explanation of symbols]
10 Measurement object 12 Measurement table 14, 16 Slit light source 18A, 18B, 18C, 18D Mark 20 Camera 20A Main camera position 20B Auxiliary camera position 24 Personal computer (PC)

Claims (2)

A first imaging unit that captures an image of the measurement target irradiated with the slit light, and captures line light that appears in a line shape on the surface of the measurement target by the slit light as a line image;
A plane for each line image on the measurement surface of the measurement object based on each line image obtained by the first imaging means each time the measurement object relatively moves by a predetermined distance with respect to the first imaging means. Three-dimensional calculating a coordinate, and calculating three-dimensional data of the surface of the measurement object based on each movement position of the measurement object detected by the position detection means and the plane coordinates calculated for each line image; Data calculation means,
A second imaging unit configured to capture the measurement target from its front position and obtain a center projection image;
A third imaging unit that captures the measurement target at a lateral position at a fixed interval from the front of the measurement target to obtain a center projection image,
A pseudo center projected image similar to the center projected image acquired by the second imaging unit is created based on the calculated three-dimensional data of the measurement target and imaging information such as the position and orientation of the first imaging unit. Means to
Means for comparing the simulated center projected image with the center projected image to determine a portion where data of the simulated center projected image is missing and should be interpolated;
The three-dimensional shape of the part to be interpolated is calculated based on the center projection image acquired by the second imaging unit, the center projection image photographed by the third imaging unit, and photographing information such as the position and orientation of each imaging unit. And a calculation means for interpolating the part. The color information of the center projection image acquired by the second imaging unit is used as the color information of the pseudo center projection image for which interpolation has been completed, and the center projection color information is projected on the three-dimensional shape data projecting the pseudo center projection image. The non-contact type shape measuring apparatus according to claim 1, further comprising a providing unit.

Images