JP2002013921A - Three-dimensional imaging device and three-dimensional imaging method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a three-dimensional imaging device for efficiently carrying out association processing in a distance information calculating configuration by color coding. SOLUTION: In three-dimensional shape measurement processing for obtaining distance data on a measuring object by irradiating it with color-coded pattern light as a projected pattern and associating the projected pattern with a photographed pattern, when the photographed pattern includes an area difficult to associate with the projected pattern, a second projected pattern is produced for irradiation by re-coding it with colors complementary to those of the first projected pattern, thereby performing association processing. By this configuration, the area where the projected pattern is difficult to associate with the imaged pattern can be efficiently eliminated and a highly accurate three- dimensional image can be obtained by relatively fewer times of imaging.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image capturing method and apparatus for obtaining distance information as well as luminance information (referred to as a three-dimensional image). The present invention relates to a three-dimensional image capturing apparatus and a three-dimensional image capturing method for realizing acquisition of a distance to a measurement target without performing the method.

[0002]

2. Description of the Related Art As a method for measuring the shape of a measurement object, a passive method (stereo image method, shape from X) is used.
And active methods (time of flight measurement method, triangulation method: spot light or pattern light projection method). Active methods include (1) a flight time measurement method that emits laser light or ultrasonic waves to measure the arrival time from an object and extract depth information, and (2) a spot light placed at a predetermined position. Using a special light source such as a slit light or the like and an imaging camera, a spot or pattern light projection method that calculates the distance based on the principle of triangulation, or (3) optical processing to form contour lines with moiré fringes and three-dimensional There is a method of obtaining information, and a method of measuring a shape by radiating some energy to a target object and detecting its reflection. There is a merit of high measurement accuracy, and therefore, various practical researches are being actively conducted. The problem with the active method is the reduction of measurement time. To solve this problem, various pattern light projection methods have been proposed from the spot light projection method, and the representative methods are the spatial coding method and the color coding method. Is the law.

As an example of the spatial coding method, Japanese Patent Laid-Open No. 5-332
This will be described with reference to an embodiment disclosed in Japanese Patent Publication No. 737. In this example, a lens system that shapes a laser light source and a laser beam into a slit shape, a scanning device that scans and irradiates the shaped laser beam and an object, a camera that detects reflected light from the object, and a It consists of a controlling device.
A stripe pattern is formed on a target object by a laser beam scanned from a scanning device at a portion irradiated with the laser beam and a portion not irradiated with the laser beam. By irradiating the laser light with a plurality of different patterns (N patterns), the target object is divided (coded) into ^2N- 1 identifiable portions. The shape of the target object can be calculated by determining which of the divided portions each pixel on the image of the target object is captured by the camera from a different position.

In general, however, in order to increase the resolution, it is necessary to perform scanning with a laser a plurality of times and perform photographing with a camera a plurality of times. For example, in order to divide the screen into 255 areas, eight times of imaging are required. For this reason, it is difficult to photograph a fast-moving object, and it is necessary to securely fix the photographing system during scanning. Therefore, even if the apparatus itself is simple, it is difficult to photograph easily.

[0005] As a second conventional example, there is a color coding method for the purpose of reducing the number of times of projecting a pattern and further shortening the measurement time. This will be described with reference to an embodiment disclosed in Japanese Patent Application Laid-Open No. 3-192474. The pattern light is encoded so that two adjacent pattern lights do not have the same color, the color of the pattern light is detected from the observed image, and the pattern light number is obtained from the color arrangement of the corresponding pattern light. The irradiation direction of the pattern light is calculated from the pattern light number, and the distance can be calculated in the same manner as in the example of space coding.

However, in the above-described method, under limited conditions, for example, when the color of the object to be measured is only white, the colors of the imaging pattern and the light emitting pattern match, and the correspondence between the two is easy. Under the environment, the measurement target has a plurality of colors (the spectral reflectance distribution is deviated), so that the imaging pattern is changed by the reflectance of the target, and it is difficult to associate the two with each other.

The spectral distribution of the light projection pattern, the spectral reflectance of the object to be measured, and the resulting reflected light will be described with reference to FIGS. FIG. 6 shows a case where all the objects to be measured are on a white background, and FIG. 7 shows a case where they are not on a white background.

As shown in FIG. 6, when the object to be measured is a white background, the object to be measured has a uniform spectral reflectance r ₀ (λ) over the entire wavelength region. For example, this measurement object
If the projection of the red projection pattern R, the reflected light quantity I _r of the projection pattern R on the measurement object,

[0009]

(Equation 1)

[0010] Here, r _o (λ): the spectral reflectance of the object to be measured i _r (λ): the spectral distribution of the light projection pattern R.

[0011] shown in FIG quantity of reflected light I _r, overlapping spectral distribution of the spectral reflectance and projection pattern R of the measurement object 6
The light amount corresponding to the gray area shown in FIG. In this case, a sufficient amount of reflected light is obtained, and the red light projection pattern R is recognized as a red imaging pattern R by the imaging device. Similarly, when the green G and blue B light projection patterns are projected, reflected light in the green and blue wavelength regions is obtained, and is recognized as a green imaging pattern and a blue imaging pattern.

However, it is generally unlikely that an object to be measured in three-dimensional measurement is a single color on a white background. FIG. 7 shows a study on the reflected light of the measurement object with respect to the light projection pattern in this case.

FIG. 7 shows the spectral distribution of a light projection pattern when the object to be measured has various color distributions (reflectance distributions) other than a white background, the spectral reflectance of the object to be measured, and the resulting reflected light. FIG. 7 (a) shows a measurement object having a distribution of reflectance r _o (λ) which is deviated to the red R region, and FIG. 7 (b)
Indicates the spectral reflectance distribution of the measurement object having a distribution of the reflectance r _o (λ) biased from the blue B to the green G region. Reflected light I _r of the projection pattern R, in turn,

[0014]

(Equation 2)

## EQU1 ## Object as shown in FIG. 7 (a), when the measuring object having a reflectance distribution leaning to red R region r _o (lambda), the amount of reflected light I _r is gray FIGS. 7 (a) Since the imaging device has a certain value as shown in the area, the imaging apparatus recognizes the pattern as a red imaging pattern R. However, as shown in FIG. for r _o measuring object having a distribution of (lambda), the amount of reflected light I _r is the reduced extremely becomes an amount corresponding to the gray area in FIG. 7 (b), be recognized as a red imaging pattern R Can not.

FIG. 7 shows an example in which a red R light projection pattern is used. However, even when green G and blue B light projection patterns are projected, the light reflection pattern of the target object is similarly changed. Therefore, there is a case where the image is not recognized as an imaging pattern of the same color.

Therefore, the color coding method has a problem that although high accuracy can be obtained in a situation where the target object is limited, the accuracy is extremely deteriorated in a general shooting situation where the target is not limited.

[0018]

SUMMARY OF THE INVENTION An object of the present invention is to solve the above-mentioned problems in the color coding method, and to solve the problem even if the spectral reflectance of the object to be measured is not uniform. It is another object of the present invention to provide a three-dimensional image capturing method and apparatus capable of accurately associating a light projection pattern with an imaging pattern and acquiring a high-accuracy three-dimensional image with a relatively small number of image capturing operations.

[0019]

SUMMARY OF THE INVENTION The present invention achieves the above-mentioned object, and a first aspect of the present invention is a projecting means for projecting a color-coded light projection pattern onto a measurement object. An imaging unit that captures an optical image generated on the surface of the measurement object by the projection unit; a comparison correspondence unit that compares and associates the imaging pattern acquired by the imaging unit with the light projection pattern; and the comparison correspondence unit. A distance calculating means for calculating a distance from the output of the light source to the object to be measured, and a light beam re-encoded by using a color having a complementary color relationship to a color used in color coding of the light emitting pattern. And a pattern replacement means for supplying a pattern.

Further, in one embodiment of the three-dimensional image pickup device of the present invention, the pattern replacement means uses a color complementary to the color used in the color coding of the light projection pattern. And a calculation processing means for performing pattern generation processing for generating a re-encoded light projection pattern.

Further, in one embodiment of the three-dimensional image pickup apparatus of the present invention, the projection pattern projected by the projection means has a k-slit configuration in which two adjacent patterns do not have the same color. It is characterized in that the arrangement of the adjacent colors does not overlap in the k slits.

Further, in one embodiment of the three-dimensional image pickup device of the present invention, the projection pattern projected by the projection means is composed of three colors of red (R), green (G) and blue (B). Projecting the projected light pattern on the object to be measured, and the pattern replacing means changes the red (R) of the projected pattern to a color having both wavelengths of green (G) and blue (B),
Green (G) is changed to a color having both red (R) and blue (B) wavelengths, and blue (B) is changed to a color having both red (R) and green (G) wavelengths. And supplying a re-encoded light projection pattern.

Furthermore, a second aspect of the present invention is a projection step of projecting a color-coded light projection pattern onto a measurement target, and an imaging step of capturing an optical image generated on the measurement target surface in the projection step. And a comparison correspondence step of comparing and associating the imaging pattern captured in the imaging step with the light projection pattern, and a color of the light projection pattern when an area that is difficult to associate is detected in the comparison correspondence step. A pattern replacement step of supplying a light-encoding pattern re-encoded by using a color having a complementary color relationship with respect to the color used for encoding; and A comparison correspondence reprocessing step of executing a correspondence processing for the re-encoded imaging pattern projected and photographed on the comparison correspondence reprocessing step. In 3-dimensional imaging method characterized in that it comprises a distance calculation step of calculating a distance to the object of measurement by the corresponding data obtained, the.

Further, in one embodiment of the three-dimensional image capturing method according to the present invention, the pattern replacement step includes:
A step of performing a pattern generation process of performing re-encoding using a color having a complementary color relationship on a color used in color encoding of the light projection pattern and generating a re-encoded light projection pattern It is characterized by including.

Further, in one embodiment of the three-dimensional image pickup method of the present invention, the projecting step measures a light projection pattern composed of three colors of red (R), green (G) and blue (B). Projecting onto a target object, wherein the pattern exchanging step comprises:
The color is changed to a color having two wavelengths of green (G) and blue (B), and the color of green (G) is changed to a color having two wavelengths of red (R) and blue (B). B) is a step of changing the color of B) to a color having two wavelengths of red (R) and green (G) and supplying a re-encoded light projection pattern.

[0026]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be specifically described below with reference to the drawings. FIG. 1 shows a configuration diagram of an embodiment of the present invention. The three-dimensional image capturing apparatus according to the present embodiment includes a projection unit 10 that projects a color-coded light projection pattern onto an object.
An imaging unit 20 that captures an optical image generated on the surface of the measurement target 100 by the projection unit 10; a comparison correspondence unit 30 that compares and associates the imaging pattern acquired by the imaging unit 20 with the light projection pattern; The distance calculation unit 40 calculates a distance to the object 100 from the output from the object 30.

In this embodiment, a liquid crystal projector 101 is used as the light source as the projection unit 10. LCD projector 10
2 is controlled by the personal computer 101 to select the color of the light projection pattern. An optical image on the measurement target 100 was captured using the CCD camera 201 as the imaging unit 20. The image output data (imaging pattern) captured by the CCD camera 201 is transmitted to the comparison correspondence unit 30 that performs comparison and association between the projection pattern data projected by the projection unit 10 and the imaging pattern data captured by the imaging unit 20. The distance calculation unit 40 calculates the distance to the measurement target 100 based on the data transmitted and obtained by the comparison corresponding unit 30.

Next, a specific color encoding method using the three-dimensional image photographing apparatus shown in FIG. 1 and a principle of distance calculation will be described.

The projection pattern was identified by changing the color of the pattern light, and the shape was a vertical stripe pattern. When the color of the pattern light is q kinds (gradation), the adjacent pattern light does not have the same gradation, and the arrangement of the gradation by k adjacent pattern light is only once for any pattern light arrangement. Given conditions that do not appear,
The number N of the total pattern lights is as shown in the following equation. That is,

[0030]

N = q (q-1) ^(k-1) + k-1

In this embodiment, the colors of the pattern light are red, green,
Color coding was performed using blue (R, G, B) with q = 3 and the number of adjacent pattern lights: k = 8. In this case, the total number of patterns N is obtained as follows by applying q = 3 and k = 8 to the above equation.

[0032]

## EQU4 ## N = 391

Next, the object to be measured is divided into 391 areas by the light projection pattern.

Next, the principle of distance calculation will be described. In Figure 2,
2 shows a layout of a light source, a camera, and an object in the present embodiment. When the pattern light is photographed by the camera, the measurement point on the captured image is P (x, y), the focal length of the camera is F, and the base line length that is the distance between the light source and the camera is L, the distance of the point P
Z is

[0035]

(Equation 5)

## EQU3 ## In the present invention, although not shown in FIG. 2, there are a plurality of light projection patterns and imaging patterns. Therefore, it is necessary to correspond to a certain imaging pattern among a plurality of projection patterns projected from the light source, and the accuracy of the correspondence greatly affects the measurement accuracy. Therefore, the following measurement method was implemented.

(Procedure 1) Color Coding by RGB Correspondence by Light Projection Pattern Using the three colors of RGB, the projection pattern is encoded by the personal computer of the projection unit 10 shown in FIG. This is performed at 101. The encoded light projection pattern is projected onto the measurement object 100 by the liquid crystal projector 102, and an optical image generated on the object surface by the light projection pattern is captured by the CCD camera 201 of the imaging unit 20.

Next, in the comparison corresponding section 30, the image pickup section 2
The imaging pattern acquired by the CCD camera 201 of No. 0 is compared with the light projection pattern projected by the projection unit 10 to execute the association processing. In an ideal state, the color data of all the imaging patterns can be obtained accurately, and the correspondence between the imaging patterns and the light projection patterns is possible. However, in reality, the color data as shown in FIG. not obtained UK pattern (u n k nown pattern) appears.

The area in which these correspondence becomes difficult, that is, the processing for UK pattern (u n k nown pattern) will be described below.

(Procedure 2) Re-encoding by Complementary Colors of RGB Correspondence by Light Projection Pattern Using the colors that are complementary colors to the light projection patterns RGB used in the color encoding of Procedure 1 for each pattern color A re-encoding process is performed. That is, R → (B + G), G → (B +
R), B → (G + R), that is, red (R) of the light emission pattern
Is changed to a color having two wavelengths of green (G) and blue (B), and green (G) is changed to a color having both wavelengths of red (R) and blue (B). Blue (B), red (R) and green (G)
Is changed to a color having both wavelengths of the two colors, color encoding is performed, a re-encoded light projection pattern is generated in the personal computer 101 of the projection unit 10, and the re-encoded light projection pattern is converted by the liquid crystal projector 102. Measurement object 10
The CCD camera 201 of the imaging unit 20 again captures an optical image generated on the surface of the object by the light projection pattern.

Note that, as shown in FIG.
It is known that before and after is a green G, and there is no pattern array of the same color continuously in the light projection pattern. Therefore, it is determined that the UK pattern is either R or B. Hereinafter, cases 1 (determined as R) and case 2 (B) for determining whether the UK pattern is R or B
Is determined).

Case 1) When the UK pattern is R When the R is not detected in the imaging pattern, the spectral reflectance of the object is roughly classified into four types as shown in FIG. 4A). If the UK pattern is R (= the first light projection pattern is R), the light projection pattern at the time of re-encoding is a color (B + G) that is complementary to the R color. Therefore, the combination of the spectral reflectance of the object and the spectral distribution of the light projection pattern can be roughly classified as shown in FIG. 4A).
There are four types shown in FIG.

FIG. 4A shows (1) that the spectral reflectance of the object to be measured is reduced only in the region of R, and (2) shows that the spectral reflectance of the object to be measured is in the regions of G and R. (3) indicates that the spectral reflectance of the object to be measured decreases in the B and R regions, and (4) indicates that the spectral reflectance of the object to be measured is all B, G, and R. It is a case where it falls in the region of FIG.

In the case where the spectral reflectance of the object to be measured is as shown in (1), (2), and (3) of FIG.
The color data of the imaging pattern imaged by the CCD camera 201 is obtained from the overlapping portion of the spectral reflectance of the measurement target and the spectral distribution of the light projection pattern.

That is, the spectral reflectance of the object to be measured is as shown in FIG.
In the case of (1) of a), a mixed color of B and G (B + G),
In the case of (2), B is strong and G is weak (B + g), (3)
In this case, B is weak and G is strong (b + G). In any case, the color data of the imaging pattern in Case 1 is a color in which B and G are mixed. Therefore, in the area (UK pattern) where the first light projection pattern and the imaging pattern cannot be associated with each other, by using the re-encoded light projection pattern using the complementary color of RGB, the complementary color is formed on the imaging pattern. A mixed color of B and G, which is similar to the light projection pattern, is reproduced, and the correspondence is possible.

The spectral reflectance of the object to be measured is shown in FIG.
As shown in (4), when the image data is reduced in all the B, G, and R regions, the color data of the imaging pattern is determined as UK again. The process in this case will be described in Procedure 3.

Case 2) Case where the UK pattern is B FIG. 4B) shows the spectral reflectance and light projection of the object when the UK pattern included in the association between the first light projection pattern and the imaging pattern is B. The combinations of the spectral distributions of the patterns are roughly divided into four as shown in FIG. UK
When the pattern is B (= first light projection pattern is B), the light projection pattern at the time of re-encoding uses a color (G + R) that is complementary to B color. Therefore, the combinations of the spectral reflectance of the target object and the spectral distribution of the projection pattern are roughly classified into four types shown in FIG. 4B).

FIG. 4B) (1) shows the case where the spectral reflectance of the object to be measured is reduced only in the region B, and (2) shows the region where the spectral reflectance of the object to be measured is B and G. (3) indicates that the spectral reflectance of the object to be measured decreases in the B and R regions, and (4) indicates that the spectral reflectance of the object to be measured is all B, G, and R. It is a case where it falls in the region of FIG.

When the spectral reflectance of the object to be measured is as shown in (1), (2), and (3) of FIG.
The color data of the imaging pattern imaged by the CCD camera 201 is obtained from the overlapping portion of the spectral reflectance of the measurement target and the spectral distribution of the light projection pattern.

That is, the spectral reflectance of the object to be measured is as shown in FIG.
b) In the case of (1), a mixed color of R and G (R + G),
In the case of (2), R is strong and G is weak (R + g), (3)
In the case of (1), R is weak and G is strong (r + G). In any case, the color data of the imaging pattern in case 1 is a mixed color of R and G. Therefore, in the area (UK pattern) where the first light projection pattern and the imaging pattern cannot be associated with each other, by using the re-encoded light projection pattern using the complementary color of RGB, the complementary color is formed on the imaging pattern. The same mixed color of R and G as in the light projection pattern is reproduced, and the correspondence is possible.

The spectral reflectance of the object to be measured is shown in FIG.
As shown in (4), when the image data is reduced in all the B, G, and R regions, the color data of the imaging pattern is determined as UK again. The process in this case will be described in Procedure 3.

As described above, the area (U) where the first projection pattern and the imaging pattern cannot be correlated.
K pattern), by using a re-encoded projection pattern using complementary colors of RGB, whether the UK pattern is originally R or B, the projection pattern is obtained by pattern projection by re-encoding using complementary colors. And the imaging pattern can be associated with each other.

As described with reference to FIG. 4, the pattern of the UK pattern is limited to Case 1 or Case 2, and the color data of the imaging pattern obtained by imaging the second projection pattern using the complementary color is: If B and G are mixed,
If the first light emission pattern is R and the color data of the imaging pattern obtained by imaging the second light emission pattern using the complementary color is a mixture of R and G, the first light emission pattern is B Can be recognized. Therefore, in the procedure 1, a UK pattern for which color data cannot be obtained accurately can be complemented.

(Procedure 3) Complementary means when color data cannot be obtained In both case 1 and case 2 of FIG. 4, the spectral reflectance of the object to be measured is reduced in all the B, G, and R regions. In such a case, no sufficient color data can be obtained by any correction means. Therefore, a UK pattern is used, and color data is inferred from the context of the pattern. The light projection pattern projected by the projection unit has a k-slit configuration in which two adjacent patterns do not have the same color, and the arrangement of the adjacent colors forming the slit does not overlap in the k-slit configuration. With this configuration, it is possible to infer from other photographing patterns.

(Procedure 4) Calculation of Distance Data Since the correspondence between the light projection pattern and the imaging pattern is determined by the procedures 1, 2, and 3, the distance is measured over the entire pattern light irradiation area by the equation (1). I do.

FIG. 5 shows the processing flow of the three-dimensional image pickup apparatus of the present invention described above. Hereinafter, each step of the processing flow of FIG. 5 will be described.

First, in step S501, the personal computer of the projection unit 10 shown in FIG. 1 generates a first projection pattern to be projected on a measurement object, projects the projection pattern by the liquid crystal projector 102, and projects the pattern projection image. Is captured by the CCD camera 201 of the imaging unit 20,
The comparison correspondence unit 30 executes a process of associating the projection pattern with the imaging pattern.

Next, in step S502, it is determined whether or not an area that is difficult to associate in the imaging pattern, that is, a UK pattern, has been detected during the association processing in the comparison association unit 30. If no UK pattern is detected, all the imaging patterns can be associated, and the process proceeds to step S506, where the distance calculation unit 40 calculates distance data based on the association processing result.

In step S502, the comparison corresponding unit 3
If a region that is difficult to associate in the imaging pattern, that is, a UK pattern, is detected in the association process at 0, the process proceeds to step S503.

In step S503, a process of associating an imaging pattern by irradiating a re-encoded pattern with a complementary color of the first projection pattern is executed. That is, as described above, for example, when the first light projection pattern uses RGB, re-encoding processing is performed using colors that are complementary to each pattern color, that is, R → (B + G), G →
(B + R), B → (G + R), color encoding is performed, a re-encoded light projection pattern is generated in the personal computer 101 of the projection unit 10, and the re-encoded light projection pattern is converted to a liquid crystal. An optical image generated on the surface of the target object by the projector 102 is projected by the projector 102, and an optical image generated on the surface of the target object by the light projection pattern is again photographed by the CCD camera 201 of the imaging unit 20. Perform attachment processing.

Further, in step S504, it is determined whether or not an area which is difficult to associate in the image pickup pattern, that is, a UK pattern has been detected during the association processing in the comparison association section 30. If no UK pattern is detected, all the imaging patterns can be associated, and the process proceeds to step S506, where the distance calculation unit 40 calculates distance data based on the association processing result.

In step S504, the comparison corresponding unit 3
If a region that is difficult to associate in the imaging pattern, that is, a UK pattern, is detected in the association process at 0, the process proceeds to step S505.

In step S505, since the spectral reflectance of the object to be measured is reduced in all the B, G, and R regions in both of the cases 1 and 2 of FIG. Even if the light projection pattern is changed, sufficient color data cannot be obtained. Therefore, a UK pattern is used, and color data is inferred from the context of the pattern.

In the above processing, the area that is difficult to associate in the imaging pattern, that is, the UK pattern is eliminated, and the process proceeds to step S506, where the distance calculation unit 40 calculates the distance data based on the result of the association processing.

In the above description with reference to FIG. 4, the spectral reflectance of the object is roughly modeled into RGB regions, but actually, the spectral reflectance exists over each wavelength region. . However, in this case, for example, in the case of FIG. 4A), the color data of the imaging pattern remains a mixed color of B and G, so that the accuracy of the association is not reduced. It should be noted that a measurement target having a sufficient spectral reflectance in the R region has the imaging pattern recognized as R in step 1 described above, and thus does not need to be considered in the modeling of the spectral reflectance in step 2.

In this embodiment, a liquid crystal projector is used as a light source, but an LED, LD, or the like may be used as another light source. At this time, in order to perform color coding using an LED or an LD, the spot light is formed into a pattern light using a rod lens or a cylindrical lens that prepares an LED or an LD according to the color used for the color coding. A predetermined pattern is collectively irradiated on the object by a scanning mechanism that scans the object with the pattern light and a control mechanism that controls the ON / OFF of the LED and LD. To do this.

In this embodiment, the colors used in the color encoding of the light projection pattern are re-encoded using complementary colors, and the re-encoded light projection pattern is re-encoded. Although the personal computer (PC) is used as the means for executing the pattern generation processing for generating the data, the present invention is not limited to the personal computer (PC), and may be any arithmetic processing means capable of performing the same processing. Further, instead of performing the complementary color-related color generation processing sequentially, a complementary color-related color pattern is prepared in advance, and if necessary, a complementary color is used for the color used in the color encoding of the light projection pattern. May be configured as a pattern exchange unit that supplies a light projection pattern re-encoded using the colors having the relationship of (1).

The present invention has been described in detail with reference to the specific embodiments. However, it is obvious that those skilled in the art can modify or substitute the embodiment without departing from the scope of the present invention. That is, the present invention has been disclosed by way of example, and should not be construed as limiting. In order to determine the gist of the present invention, the claims described at the beginning should be considered.

[0069]

As described above, according to the three-dimensional image capturing apparatus and the three-dimensional image capturing method of the present invention, a pattern light coded by color is irradiated as a projection pattern, and the light projection pattern and the imaging pattern In the three-dimensional shape measurement processing for acquiring the distance data of the measurement target by associating, if an area that is difficult to associate with the imaging pattern is included,
Since the re-encoding process is performed with a color that is complementary to the first light projection pattern, a second light projection pattern is generated and irradiated, and the association process is performed. It is possible to efficiently eliminate an area in which it is difficult to make the correspondence, and acquire a highly accurate three-dimensional image with a relatively small number of times of imaging.

Further, according to the three-dimensional image pickup apparatus and the three-dimensional image pickup method of the present invention, the spectral reflectance of the object to be measured is not uniform, and the image pickup pattern obtained by irradiating the light projection pattern has Even when a region that is difficult to associate is included, the region that is difficult to associate can be efficiently eliminated by using a new light projection pattern obtained by re-encoding the light projection pattern.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration of a three-dimensional image capturing apparatus according to the present invention.

FIG. 2 is a diagram showing a distance calculation method by a spatial coding method of a three-dimensional shape measurement device used in the three-dimensional image pickup device of the present invention.

FIG. 3 is a diagram illustrating a difficult-to-associate area (UK pattern) in an imaging pattern acquired by an imaging unit in the three-dimensional imaging apparatus of the present invention.

FIG. 4 is a diagram illustrating a process of eliminating a difficult-to-associate area (UK pattern) in an imaging pattern using a complementary color in the three-dimensional image imaging apparatus of the present invention.

FIG. 5 is a processing flowchart showing processing in the three-dimensional image capturing apparatus of the present invention.

FIG. 6 is a diagram illustrating a spectral distribution of a light projection pattern, a spectral reflectance of an object to be measured, and reflected light obtained as a result (when all the objects to be measured are white backgrounds).

FIG. 7 is a diagram illustrating a spectral distribution of a light projection pattern, a spectral reflectance of an object to be measured, and reflected light obtained as a result (when the object to be measured is not a white background).

[Explanation of symbols]

 REFERENCE SIGNS LIST 100 measurement object 10 projection unit 101 personal computer 102 liquid crystal projector 20 imaging unit 201 CCD camera 30 comparison corresponding unit 40 distance calculation unit 301 UK pattern

Claims (7)

[Claims] A projection means for projecting a color-coded light projection pattern onto a measurement target; an imaging means for capturing an optical image generated on the surface of the measurement target by the projection means; A comparison correspondence unit that compares and associates an imaging pattern with the light projection pattern, a distance calculation unit that calculates a distance to a measurement target from an output from the comparison correspondence unit, and a color coding of the light projection pattern. Pattern replacement means for supplying a light-projection pattern re-encoded by using a color having a complementary color relationship to the used color. 2. The pattern exchanging means performs re-encoding using a color having a complementary color relationship with a color used in color encoding of a light-projected pattern. 2. The three-dimensional image pickup apparatus according to claim 1, wherein the three-dimensional image pickup apparatus has a configuration including an arithmetic processing unit that executes a pattern generation process for generating a pattern. 3. A projection pattern projected by the projection means,
The k-slit configuration in which two adjacent patterns do not have the same color, and the arrangement of adjacent colors forming the slit is a configuration that does not overlap in the k-slit configuration. 3. The three-dimensional image capturing device according to 1. 4. A projection pattern projected by the projection means,
A light projection pattern composed of three colors of red (R), green (G), and blue (B) is projected on the object to be measured, and the pattern replacement means converts the red (R) of the light projection pattern into green (G) is changed to a color having two wavelengths of blue (B), and green (G) is changed to a color having two wavelengths of red (R) and blue (B).
2. The structure according to claim 1, wherein blue (B) is changed to a color having two wavelengths of red (R) and green (G) to supply a re-encoded light projection pattern. A three-dimensional image capturing apparatus according to claim 1. 5. A projecting step of projecting a color-coded light projection pattern onto a measurement target; an imaging step of capturing an optical image generated on the surface of the measurement target in the projection step; A comparison corresponding step of comparing and associating an imaging pattern with the light projection pattern, and, when an area that is difficult to associate is detected in the comparison corresponding step, a color used in color coding the light projection pattern. On the other hand, a pattern replacement step of supplying a light-projected pattern re-encoded by using a color having a complementary color relationship, and a re-encoding that is performed by projecting the re-encoded light projecting pattern onto a measurement object and photographing the same. A comparison correspondence reprocessing step of executing a correspondence processing for the imaging pattern, and a correspondence data obtained in the comparison correspondence reprocessing step. 3D imaging method characterized in that it comprises a distance calculation step of calculating the distance to the measurement object, the. 6. The pattern exchanging step includes re-encoding a color used in the color encoding of the light projection pattern using a color having a complementary color relationship, and 6. The method according to claim 5, further comprising a step of executing a pattern generation process for generating a pattern.
Dimensional image capturing method. 7. The projecting step is a step of projecting a light projection pattern composed of three colors of red (R), green (G), and blue (B) onto an object to be measured. Changing the red (R) of the light emitting pattern to a color having both wavelengths of green (G) and blue (B), and changing green (G) to two colors of red (R) and blue (B). Change to a color that has the color wavelength,
6. The method according to claim 5, wherein blue (B) is changed to a color having two wavelengths of red (R) and green (G) to supply a re-encoded light projection pattern. The three-dimensional image capturing method according to the above.