JPH03293507A - Three-dimensional shape measuring apparatus - Google Patents

Abstract

PURPOSE:To improve the measuring accuracy of the shape or the like of an object by making it possible to detect the phase of each pattern image with use of a plurality of kinds of code patterns, thereby to measure the shape while using a video image of the pattern image obtained only by projecting and photographing a bright and dark pattern as a reference. CONSTITUTION:A code pattern switching means switches sequentially a group of code patterns, so that each code pattern is projected on the surface of an object to be measured. A phase detecting means photographs an image of each code pattern projected to the object, and decodes the image of a code based on the presence or absence of a slit image at a predetermined position while all of the images of the code patterns are layered. The code is an independent one in accordance with the difference of the phase. The code is detected as a phase of the pattern image belonging to the predetermined position. When the phase of each pattern image is detected, the pattern image is compared with a bright and dark pattern of the same phase at a light source, so that the shape or position of the object can be measured according to the trigonometrical survey. In other words, it is possible to measure the shape while an image of the bright and dark pattern is used as a reference.

Description

Detailed Description of the Invention ■ Leg Shell Cup [Industrial Application Field] The present invention relates to a three-dimensional shape measuring device, and more specifically, it measures the shape of an object from an image captured by irradiating a bright and dark pattern from a light source onto the surface of the object to be measured. The present invention relates to a three-dimensional shape measuring device that measures the shape and position of a measurement object.

[Prior Art] When measuring the shape of an object to be measured from a line image obtained by irradiating, for example, striped light from a light source in a bright and dark pattern on the surface of the object to be measured, each line image on the image is Correspondence with each slit on the light source side that formed the line image,
That is, this type of shape measurement using the trigonometric method is not possible unless the phase of each line image is detected. For example, if the surface of the object to be measured is not a smooth free-form surface, a striped image will appear on the image, making it impossible to determine which slit on the light source side the line image corresponds to, making it impossible to measure the shape. Sometimes.

Therefore, a three-dimensional shape measuring device has been proposed, for example, as described in Japanese Patent Application Laid-Open No. 60-152903. This device irradiates multiple types of light patterns using multiple types of 1 mm slit patterns, as described below, and processes the resulting multiple images in combination to create a line image with a clear phase. It combines images and enables shape measurement using triangulation.

In other words, when multiple types of slit patterns used in this device (all slit patterns are stacked) and the presence or absence of a slit at a predetermined position is replaced with a binary code, independent codes can be obtained depending on the position. Accordingly, the entire slit pattern area is formed so as to be divided into a plurality of striped areas.

Therefore, if these slit patterns are sequentially irradiated onto the surface of the object to be measured and imaged, multiple images with different slit image shapes and arrangements will be obtained, but if the luminance data of these images is binarized for each pixel at the same coordinates, , the above binary code is obtained for each coordinate from the arrangement (arrangement of presence/absence of slit images). Then, the pixels are classified based on the difference in binary code, and a line image obtained by dividing the screen into a plurality of striped areas is synthesized.

In the image of this synthesized line image (since the phase of the horizontal line image is clear from the above binary code, the correspondence between each line image on the image and the striped area divided at the slit pattern position on the light source side is established). As mentioned above, even if the line image is cut, it is possible to measure the shape using trigonometry.

[Problem to be solved by the invention] However, the above-mentioned conventional device may cause erroneous measurements;
The problem is that high measurement accuracy cannot be expected.

In other words, because conventional equipment synthesizes a line image from multiple images, if the light source that projects the slit pattern, the measurement optical system such as the camera, or the object to be measured moves even slightly during imaging, the image on the screen may change. In this case, the line image is formed in the wrong shape, and the object to be measured is measured in the wrong shape.

The three-dimensional shape measuring device of the present invention aims to solve the above problems and improve the accuracy of this type of measurement.

Structure of Soft Skin [Means for Solving the Problem] As illustrated in FIG. A three-dimensional shape measurement in which the shape or position of the object to be measured is determined by detecting the phase of each bright and dark pattern image from an image of the bright and dark pattern image obtained by capturing the bright and dark pattern image with an imaging means. In the apparatus, when the presence or absence of a slit of a predetermined phase is encoded in a state where all of the patterns are stacked and are made up of multiple types of slit patterns, a group of code patterns that constitute independent codes due to the difference in phase, and the code code pattern switching means for sequentially switching each code pattern of the pattern group in front of the light source; The apparatus is characterized by comprising a phase detection means for decoding the code based on the presence or absence of a slit image and detecting the code as the phase of each of the pattern images belonging to the predetermined position.

[Operation] Using the three-dimensional shape measuring device of the present invention having the above configuration, the phase of each bright and dark pattern image is calculated from the images of the bright and dark pattern images obtained by irradiating the object to be measured with light in a bright and dark pattern and capturing the image. Detect it like this.

First, the code pattern switching means 1 sequentially switches each code pattern of the code pattern group and projects each code pattern onto the surface of the object to be measured.

Next, the phase detection means captures an image of each code pattern projected onto the object to be measured, and decodes the code based on the presence or absence of a slit image at a predetermined position in a state in which all of the images are stacked. The signs take independent results due to the difference in phase,
This code is detected as the phase of each of the above pattern images belonging to a predetermined position.

When the phase of each pattern image is detected, each pattern image and
It corresponds to the light and dark pattern of the same phase in the light source,
The shape or position of the object to be measured is measured by trigonometry.

In other words, it becomes possible to measure the shape based on a bright and dark pattern image obtained by irradiating and imaging a bright and dark pattern.

[Example] A three-dimensional shape measuring device as an example of the present invention will be described below with reference to the drawings.

Three-dimensional shape measuring device (1: as shown in Figure 2).

It is composed of a projector 10, a camera 20, a signal processing device 30, and a surface shape display device 40.

The projector 10 includes a lamp 11, a condenser lens 2, a liquid crystal shutter 13, and a projection lens 14.

Liquid crystal shutter 13 (as shown in FIG. 3, transparent electrodes 16 are formed in stripes at predetermined intervals on the inner surface 1 of a 1mm polarizing glass plate 15, and transparent electrodes 18 are formed on the inner surface of the opposing polarizing glass plate 17. The image electrodes 16 and 18 are formed into a planar shape.
A liquid crystal 19 is filled in between. 1 to the transparent electrode 16. t, power is individually applied from the pattern control unit 32 as described later. As a result, the molecular arrangement direction of the liquid crystal 19 changes, and various projection patterns are created by combining the stripe pattern of the transparent part and the stripe pattern of the light-shielding part (see FIG. 4). Transparent electrode 18 (a common electrode that corresponds to all of the companion transparent electrodes 16).The liquid crystal shutter 13 of the embodiment has a light-shielding property when an electric field is applied. It may also be made translucent.

Therefore, as shown in FIG. 2, the light from the 1:S lamp 11 passes through the condenser lens 12 and then becomes, for example, striped light through the liquid crystal shutter 13, and is irradiated onto the surface of the object to be measured through the projection lens 14. Ru.

A camera 20 (as shown in FIG. 2, in this embodiment, it has an objective lens 22 and an image sensor 24, and is constituted by, for example, a normal low-speed scanning television camera or a CCD camera. Image sensor 24) An image of the surface of the object to be measured M is formed above.

The signal processing device 30 is configured as a computer tL,
It includes a CPU 30a, a ROM 30b, a RAM 30C, and an interface 30d, and is configured to be able to mutually transmit signals via a bus 30e.

Furthermore, the F A/D converter 3 is installed on the interface 30d.
1, a pattern control section 32 and a surface shape display device 40 are connected.

A/D converter 31[;& According to a command from CPU 30a, an analog signal from camera 20 is converted into a digital signal and output to interface 30d.

Pattern control unit 32 [1CPU 30a commands to supply power to the transparent electrode 16 of the liquid crystal shutter 13 to create various projection patterns that combine the stripe pattern of the light-transmitting part and the stripe pattern of the light-shielding part.

FIG. 4 shows a perspective view of an example of the projection pattern, and FIG. 5 shows an enlarged front view of a portion of the projection pattern. In the embodiment, one striped pattern SP and eight code patterns CP are created. Note that for convenience, the light-shielding portions are indicated by hatching.

The striped pattern SP is formed by alternately forming light-transmitting portions and light-blocking portions for each stripe pattern (transparent electrode 16). In the embodiment, the width Wl of one stripe pattern (transparent electrode 16) is large enough to allow a plurality of pixels to capture an image of the pattern.

On the other hand, each code pattern CPI table is a combination of a light-transmitting part and a light-blocking part with a plurality of stripe patterns as basic units. Each code pattern CP is created according to the following rules.

FIG. 6 shows 8-digit (8-bit) Gray codes (alternating binary codes) GC arranged vertically. Each code pattern CP is created by looking at the array of gray codes GO in the vertical direction for each digit and replacing the value 1 with the stripe pattern of the light-transmitting part and the value O with the stripe pattern of the light-blocking part.

Therefore, when eight code patterns CP are stacked, regarding the arrangement of overlapping transparent parts and light-blocking parts with the same set number (value corresponding to the phase), the transparent parts are decoded with the value 1 and the light-blocking parts with the value 0. Then, an independent code, in this case a Gray code GC, is obtained due to the difference in phase.

In the example (as shown in FIG. 5), each code pattern C
The switching position (transition position) of the light-transmitting/light-blocking pattern of P is aligned with the switching position (transition position) of each set of the striped pattern SP, and within the same striped area of the striped pattern SP ( The configuration is such that the same gray code GC is assigned.

Surface shape display device 40 (displays the shape of the object to be measured M finally determined through the processes described below in the form of a three-dimensional image or the like).

Next 1: The shape measurement process executed in the signal processing device 30 will be explained based on the flowchart of FIG.

The shape measurement processing routine is executed every time a measurement start command is issued after starting up the three-dimensional shape measuring device.

When this routine is started, first, the pattern of the liquid crystal shutter] 3 is switched to a striped pattern SP, and the striped pattern SP is projected onto the object to be measured M (step 100).

Next, the surface of the object to be measured M is imaged, and the video signal is converted to A.
The image data digitally converted by the /D converter 31 is
Processing to store in a predetermined area of M30c (Step 1] 0)
Do the following. FIG. 8 1: illustrates an example of a stored image. In the embodiment, one image is composed of pixels arranged in 512 rows and 512 columns. Furthermore, the illuminance data of the pixels is expressed in 256 gradations.

Next, as shown in FIG.
The 512 illuminance data groups shown in Figure 1 are read (step 120). FIG. 9 illustrates a part of the illuminance data D.

Next, the fundamental wave f is calculated by interpolation from the set of points represented by this one row of illuminance data D in 256 gradations (step 30; illustrated in FIG. 9). Since interpolation is a well-known process, details will be omitted.

After this, the peak position PP of each wave in the fundamental wave f is detected (step 140; exemplified in FIG. 9), and the peak position PP
Identify the coordinates of the pixel.

Next, it is determined whether or not the above-mentioned peak position PP detection processing has been completed for all other rows (1=0 to 511) (Step 150), and if it has not been completed (L next row (first is i If = 0, then 1 = 1 column), read 512 illuminance data groups in the Xc force direction (step 1
20), calculate the fundamental wave f by interpolation step 130
), the peak position PP of each wave in the fundamental wave f is detected (step 140).

This process is repeated until all rows (i = O~51
When the detection of the peak position PP for 1) is completed, an initial value 1 is set to a variable n that specifies the code pattern CP (step 160).

Next, the liquid crystal shutter 13 is switched to the code pattern CPI specified by the variable n=1 and projected onto the object to be measured M (step 170). Then, the surface of the object to be measured M is imaged, and the image data obtained by digitally converting the video signal by the A/D converter 31 is stored in a predetermined area of the RAM 30c (step) 80.

After this, in the stored image of the code pattern CPI, the illuminance data of the pixel at the same coordinates as the coordinates of the peak position 1PP (step 140) is binarized using a threshold value (step 1
90). As a result, the Gray code component (value) or value 0 for the peak position PP indicated by the code pattern CP1 is obtained. For example, in the example shown in FIG. 9, the gray code component of the peak position iPP indicated by the code pattern CPI has the value O (shaded image).The gray code component detection process (step 90) is performed in step 90. ]
This is performed for all peak positions obtained by repeating the process 40 times.

Next, the Gray code constituent elements located between the respective peak positions thus obtained are stored in a predetermined area of the RAM 30c, and decoding processing (step 2oO) is performed.

The decoding process here is a process of arranging the Gray code constituent elements obtained in the above process (step 190) after the Gray code constituent elements obtained in the previous process (step 190), in which the peak position Do this for each of the following.

The decoding process from code pattern CPI to code pattern CP8 is repeated to configure gray code GC.

Step 1: It is determined whether the variable n has the value 8, that is, whether the decoding process for all code patterns CP has been completed (step 210), and if it has not been completed (the value of the variable n
The value 1 is added to make the value 2 (step 220), and the process returns to step 170.

In step 170, the pattern of the upper liquid crystal shutter 13 is switched to the code pattern CP2 and an image is captured, and in the image, the illuminance data of pixels at the same coordinates as the peak position PP is binarized (step 190), and the code pattern C
The Gray code component (value 1 or value O) of each peak position related to P2 is obtained.

By repeating the above process, the Gray code components of each peak position are obtained from all the images of the eight code patterns CP (step 190), and when they are arranged and decoded (step 200), Gray code GC is obtained. In the example of FIG. 9 (table), the gray code regarding the peak position PP is converted into a set number and has a value of 0.

Finally, step 23 calculates the shape of the measured object M based on the trigonometric method from the peak position and its gray hood, etc.
0), this process ends. The calculated shape of the object to be measured M is displayed on the surface shape display device 40 as required.

Shape calculation (coordinates (xc, yc) of the peak position Pc of the line image formed on the image sensor 24 as shown in FIG. 10)
The shape of the object to be measured is determined from the set number m expressed in decimal notation of the gray code GC, and the optical system installation parameters (Ll, L, 2°11, +2.θ) of the projector 10, camera 20, etc. Representative coordinate point (×,・Y,
Z) is performed according to the following formula.

X=xc・(L2-Z)/I2Y=yc・(L2-Z)/I2A=-H-co
sθ, B = −12·cosθ.

C=-sinθ, D=-H・I2・sin
θ.

E=-HL-cO5θ, F=+2-Ll.

G=-L: L sin θ, H=2 yr I 1/S
Oφ=π·m Here, φ is the phase of the line image. The phase φ is obtained directly from the set number m. Further, SO is the pitch of the stripe pattern of one set of light-transmitting and light-blocking parts in the striped pattern SP (also shown in FIG. 5).

As explained above, the three-dimensional shape measuring device (like
Using the images of eight code patterns CP, the configuration is such that the phase of the line image of one line image related to one striped pattern SP can be checked, and the shape measurement from one line image image is possible. Therefore, even if the measurement optical system such as a projector or camera or the object to be measured moves during imaging, it is possible to obtain high measurement accuracy, which is an excellent effect.

In addition, since the phase of the line image is determined using the gray code GC in the example, for example, the coordinates of the peak position of the line image are
This corresponds to the transition point (boundary between a pixel in a light shielding area and a pixel in a transparent area) in the code pattern equivalent image (see Figure 9), and even if a gray code component is read incorrectly, the error is ±1.
(Hamming distance is 1) and small. Therefore, for example, if the surface of the object to be measured M is a smooth free-form surface, the effect of the error can be removed by a simple correction, and a high measurement accuracy can be obtained. be.

Furthermore, in the embodiment, as described above (FIG. 5), the switching position (transition position) of the light transmitting/light blocking pattern of each code pattern CP is the same as the switching position (transition position) of each stitch of the striped pattern SP. The same gray code GC is assigned to the same striped range of the striped pattern SP.

Therefore, as illustrated in FIG.
The coordinates of the pixel (thick line frame) where the peak position PP appears in the image corresponding to P frame) results in a location nearby.

Near this central pixel, contrast information (luminance data) is most stable, and accurate decoding results can be obtained.

In addition, even if the measurement optical system and the object to be measured move during imaging and each code pattern image shifts, the same decoding result as for the center pixel can be obtained, just because the peak position deviates from the position near the center pixel.

Based on these facts, the apparatus of the embodiment can accurately assign a predetermined fringe number to each line image of the line image and detect the phase of the line image correctly, making it possible to significantly improve measurement accuracy. can.

Although the embodiments of the present invention have been described above, the present invention is not limited to these embodiments, and it goes without saying that the present invention can be implemented in various forms without departing from the gist of the present invention. For example, the present invention may be applied to a measuring device that uses a dot pattern that irradiates light in a dot matrix as the light-dark pattern. If the configuration uses the example code pattern CP as is (it becomes possible to detect the phase (number) of each dot image from the captured image), it is possible to detect the phase (number) of each dot image from the captured image. Therefore, for example, by projecting and imaging one dot pattern, It becomes possible to accurately measure the shape of

Moreover, instead of the code pattern to which the Gray code is applied, a code pattern to which the normal pure binary code is applied may be used. Zara (instead of a liquid crystal shutter, PLZ
A T electro-optic shutter array may also be used. A CRT may be used as a projector to project various patterns. The projection order of the patterns does not matter. In addition, a configuration may be adopted in which the images of the striped pattern SP and the code pattern CP are taken in at the same time, and then various calculations are performed once. With this configuration, since the image capture time is short, the influence of the movement of the object to be measured becomes much smaller (2 microscopic), and the phase of the line image can be specified extremely accurately, resulting in a further improvement in measurement accuracy.

Furthermore, in the example, all eight code patterns CP are used to determine the stripe number and, therefore, the phase, but the coarsest code pattern CPI (pattern that determines the positive or negative sign of the stripe number) and the next Coarse patterns such as coarse code pattern CP2 (a configuration in which the above is omitted may also be used). This means that even if the same stripe number is given to multiple line images by omitting the code pattern, as illustrated below, the line image This is because it is possible to detect the correct phase by separating them from each other.For example, (i) As can be understood from the table of fringe numbers shown in Figure 6, 1 is a stripe pattern in which the code pattern CPI is omitted and no positive or negative sign is attached. Even if we consider the numbers, the line images corresponding to the same value of the stripe number are far apart from each other, so each line image can be identified separately.With a configuration in which some of these code patterns are omitted, the time required for imaging is shortened. Therefore, there is an advantage that the measurement time can be shortened.

Effects of the Invention As detailed above, the three-dimensional shape measuring device of the present invention makes it possible to detect the phase of each pattern image using a plurality of types of code patterns, and it is possible to detect the phase of each pattern image by simply irradiating and imaging bright and dark patterns. Since it is possible to measure the shape using the image of the pattern image as a reference, the movement of the object to be measured or the measuring device during imaging will not affect the measurement, and the accuracy of this type of measurement can be significantly improved. play.

[Brief explanation of drawings]

FIG. 1 is a block diagram illustrating the basic configuration of the present invention, FIG. 2 is a block diagram of a three-dimensional shape measuring device as an embodiment of the present invention, and FIG. 3 (A) is a front view of a liquid crystal shutter. Figure 3 (B is a cross-sectional view thereof, Figure 4 is a perspective view illustrating the projection patterns created by the liquid crystal shutter, Figure 5 is an explanatory diagram showing a part of each projection pattern enlarged, and Figure 6 is An explanatory diagram illustrating the pattern of the code pattern, FIG. 7 is a flowchart illustrating an example of shape measurement processing performed in the signal processing device, FIG. 8 is an explanatory diagram illustrating an image of a striped pattern, and FIG. 9 is a line diagram. FIG. 10 is an explanatory diagram illustrating the state of the phase-lateral ratio of the image, and FIG. 10 is an explanatory diagram illustrating the installation parameters of the optical system. 10...-Projector 20...Camera SP...Striped pattern GC... Gray code PP--Peak position 13--Liquid crystal shutter 30--Signal processing device CP--Code pattern (code)

Claims (1)

[Scope of Claims] 1. A light-dark pattern image is formed on the surface of an object to be measured by irradiating a light-dark pattern from a light source, and the light-dark pattern image is captured by an imaging means. From the image of the light-dark pattern image,
A three-dimensional shape measuring device that detects the phase of each bright and dark pattern image to determine the shape or position of the object to be measured, which consists of multiple types of slit patterns, and when all of the patterns are stacked, a slit of a predetermined phase is formed. a code pattern group that constitutes independent codes due to phase differences when encoding the presence or absence of the object; a code pattern switching means for sequentially switching each code pattern of the code pattern group in front of the light source; and the object to be measured. Each code pattern projected on the surface is imaged, and the code is decoded based on the presence or absence of the slit image at a predetermined position in a state in which all of the images are stacked, and the code is converted into each pattern image belonging to the predetermined position. A three-dimensional shape measuring device comprising: phase detection means for detecting the phase of the three-dimensional shape.