JP2016040539A - Noise detector and noise detection method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To detect noise when three-dimensionally measuring a measurement object in a pattern light projection method.SOLUTION: A noise detector comprises: a whole irradiation control part 41 controlling a projector 1 so as to irradiate light onto a whole measurement area; a non-irradiation control part 42 controlling the projector 1 so as not to irradiate light onto the whole measurement area; a threshold setting part 43 setting a threshold value for noise detection to each coordinate of both images on the basis of luminance values of both images imaged by an imaging apparatus 2 in a state where irradiation of light by the projector 1 is controlled by the whole irradiation control part 41 and the non-irradiation control part 42; a noise detection part 45 which determines whether a difference between the luminance values of both images imaged by the imaging apparatus 2 in a state where light of projection patterns and light of reverse patterns are irradiated by the projector 1 in a three-dimensional measurement is equal to or less than a threshold of a corresponding coordinate set by the threshold setting part 43, and detects the coordinate as noise when the projection patterns which are equal to or less than the threshold value are equal to or more than a prescribed number of patterns; and a noise removal part 46 removing the information of the coordinate detected by the noise detection part 45.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a noise detection apparatus and a noise detection method for detecting noise generated when a pattern light projection method is used.

  Conventionally, a three-dimensional measurement method using a pattern light projection method has been proposed (see, for example, Patent Document 1). In the method disclosed in Patent Document 1, imaging is performed in a state where a plurality of encoded projection patterns are irradiated onto a measurement object. Then, the image obtained by each projection pattern is binarized in light and dark for each pixel, and the space is encoded based on the contents of the binarized pattern. Then, triangulation is performed based on the encoding result, and the three-dimensional shape of the measurement object is calculated.

  An example of a projection pattern used in the pattern light projection method is “binary code pattern”. This binary code pattern uses a numerical representation in binary numbers. For example, when encoding three spaces with eight projection patterns, the projection pattern is as shown in FIG. Then, by setting the area irradiated with light of this projection pattern to “1” and the area not irradiated to “0”, the space can be encoded as shown in FIG.

  Further, as a method for determining whether each area of the measurement area is “1” or “0”, the projection pattern is irradiated with light of a pattern (inversion pattern) obtained by inverting the brightness of the projection pattern. A method of judging from the magnitude relationship of luminance values obtained with the inversion pattern is generally implemented. For example, if the brightness value of the target point of the image obtained with the projection pattern and the reverse pattern is acquired and “the brightness value obtained with the projection pattern> the brightness value obtained with the reverse pattern”, the target point is the light of the projection pattern. Is determined to be an illuminated point, and vice versa.

JP-A 64-0554208

On the other hand, in the pattern light projection method, since the space is encoded using the light and darkness and shading of the projection pattern irradiated to the measurement object and three-dimensional measurement is performed based on triangulation, three-dimensional measurement is performed when an error occurs in encoding. An error also occurs.
For example, when the light of the projection pattern is reflected by the uneven surface due to the unevenness of the measurement object, the reflected light may be irradiated to the region that should not have been originally irradiated. In this case, in this region, it is not possible to obtain the light and darkness and lightness and darkness of the correct projection pattern, which may cause an error in encoding.

  As a specific example, as shown in FIG. 7, a measurement object (concave workpiece 10) is arranged on a tray 11, the projector 1 is arranged obliquely above the workpiece 10, and the imaging device 2 is directly above. Consider a case where encoding is performed with a total of six binary code patterns (three projection patterns shown in FIG. 5 and their respective inversion patterns).

  In the measurement environment shown in FIG. 7, the case where the light of the 1st projection pattern of FIG. 5 and its inversion pattern is irradiated is shown in FIG. The light and darkness of the tray 11 in FIGS. 8A and 8B indicate the light irradiation area when the work 10 is not on the tray 11, and the white area is the light irradiation area, and the black area is the black area. This is a region that is not irradiated with light. In addition, FIG. 8C shows the luminance value at the point of interest (circle point in the figure) among the images picked up by the image pickup apparatus 2 in the state shown in FIGS. 8A and 8B.

  Here, when there is no work 10 on the tray 11, the luminance value at the attention point is high in the first projection pattern and low in the reverse pattern. On the other hand, when the workpiece 10 is on the tray 11, when the light of the first projection pattern is irradiated, the direct light located on the right side of the broken line in FIG. 8A is blocked by the side wall in front of the workpiece 10. However, since the reflected light from the back side wall reaches the point of interest, a high luminance value is obtained. In the reverse pattern, the luminance value is low. Therefore, as shown in FIG. 8C, the luminance value of the first projection pattern is higher than the luminance value of the inverted pattern, and it is determined that the light is irradiated. Therefore, there is no accidental problem. Is obtained. Although FIG. 8 shows a case where reflected light is strongly obtained, the magnitude of the obtained luminance value varies depending on the conditions under which the light is shielded or reflected.

  Next, FIG. 9 shows a case where light of the second projection pattern of FIG. 5 and its inverted pattern is irradiated in the measurement environment shown in FIG. Also, FIG. 9C shows the luminance value at the point of interest (circle point in the figure) among the images taken by the imaging device 2 in the state shown in FIGS. 9A and 9B.

  Here, when there is no work 10 on the tray 11, the luminance value at the point of interest is low in the second projection pattern and high in the reverse pattern. On the other hand, when the workpiece 10 is on the tray 11, when the light of the second projection pattern is irradiated, the light is not originally irradiated, but the reflected light from the back side wall reaches the point of interest, and thus a high luminance value. Is obtained. Further, even in the reverse pattern, the direct light located on the right side of the broken line in FIG. 9B is blocked by the side wall in front of the workpiece 10, but the reflected light from the back side wall reaches the point of interest, and thus has a high luminance value. Is obtained. Here, since the determination of whether or not the light of the projection pattern is irradiated is based on the magnitude of the luminance value obtained from the image, the luminance value in the projection pattern and the inversion pattern as shown in FIG. 9C. If they are almost equal, correct judgment cannot be made. Therefore, there is a risk of leading to a coding error. Although FIG. 9 shows the case where the luminance values of the projection pattern and the inversion pattern are substantially equal, the magnitude of the difference between the luminance values varies depending on the conditions under which light is shielded or reflected.

  Next, FIG. 10 shows a case where light of the third projection pattern of FIG. 5 and its inverted pattern is irradiated in the measurement environment shown in FIG. FIG. 10C shows the luminance value at the point of interest (the circle in the figure) among the images taken by the imaging device 2 in the states shown in FIGS.

  Here, when there is no work 10 on the tray 11, the luminance value at the point of interest is low in the third projection pattern and high in the reverse pattern. On the other hand, when the workpiece 10 is on the tray 11, when the light of the third projection pattern is irradiated, the light is not originally irradiated, but the reflected light from the back side wall reaches the point of interest, and thus a high luminance value. Is obtained. Further, even in the reverse pattern, the direct light located on the right side of the broken line in FIG. 10B is blocked by the side wall in front of the workpiece 10, but the reflected light from the back side wall reaches the point of interest, and thus has a high luminance value. Is obtained. Here, since it is determined whether or not the light of the projection pattern is irradiated based on the magnitude of the luminance value obtained from the image, the luminance value in the projection pattern and the inversion pattern as shown in FIG. If they are almost equal, correct judgment cannot be made. Therefore, there is a risk of leading to a coding error. Note that FIG. 10 shows a case where the luminance values of the projection pattern and the inversion pattern are substantially equal, but the magnitude of the difference between the luminance values varies depending on the conditions under which light is shielded or reflected.

  The present invention has been made to solve the above-described problems, and a noise detection apparatus and noise capable of detecting noise when performing three-dimensional measurement on a measurement object by a pattern light projection method. It aims to provide a detection method.

  The noise detection device according to the present invention includes an entire irradiation control unit that controls the projector to irradiate the entire measurement region, and a non-irradiation control unit that controls the projector so that the entire measurement region is not irradiated with light. In order to detect noise for each coordinate of the image based on the luminance value of both images captured by the imaging device in a state where the irradiation of light by the projector is controlled by the total irradiation control unit and the non-irradiation control unit The difference between the brightness values of the two images captured by the imaging device in a state in which the projection pattern and the light of the inverted pattern are irradiated by the projector in the three-dimensional measurement is set by the threshold setting unit. A noise detection unit that determines whether or not the number of projection patterns that are less than or equal to the threshold value is equal to or greater than a predetermined number, and that detects the coordinates as noise; In which noise is a noise removing unit that removes the information of the detected coordinates by parts.

  According to this invention, since it comprised as mentioned above, when performing three-dimensional measurement with respect to a measurement object by the pattern light projection method, noise can be detected.

It is a figure which shows the structure of the three-dimensional measurement detection system which concerns on Embodiment 1 of this invention. It is a figure which shows the structure of the noise detection apparatus which concerns on Embodiment 1 of this invention. It is a flowchart which shows operation | movement of the noise detection apparatus which concerns on Embodiment 1 of this invention. It is a figure explaining the threshold value setting by the noise detection apparatus which concerns on Embodiment 1 of this invention, (a) It is a figure which shows the case where light is irradiated to the whole measurement area | region, (b) Light is applied to the whole measurement area | region. It is a figure which shows the case where it does not irradiate, and is a figure which shows the luminance value in the round-dot position shown to (c) (a), (b). It is a figure which shows an example of a binary code pattern. 6 is a table showing encoding of a space using the binary code pattern shown in FIG. It is a figure which shows an example of the measurement environment of three-dimensional measurement. FIG. 8 is a diagram illustrating a case where light is irradiated using the first projection pattern illustrated in FIG. 5 in the measurement environment illustrated in FIG. 7, and (a) a diagram illustrating a case where light of the first projection pattern is irradiated. (B) It is a figure which shows the case where the light of the inversion pattern of a 1st projection pattern is irradiated, and is a figure which shows the luminance value in the round-dot position shown to (c) (a), (b). FIG. 8 is a diagram illustrating a case where light is irradiated using the second projection pattern illustrated in FIG. 5 in the measurement environment illustrated in FIG. 7, and (a) a diagram illustrating a case where light of the second projection pattern is irradiated. (B) It is a figure which shows the case where the light of the reversal pattern of a 2nd projection pattern is irradiated, and is a figure which shows the luminance value in the round dot position shown to (c) (a), (b). FIG. 8 is a diagram illustrating a case where light is irradiated using the third projection pattern illustrated in FIG. 5 in the measurement environment illustrated in FIG. 7, and (a) a diagram illustrating a case where light of the third projection pattern is irradiated. (B) It is a figure which shows the case where the light of the reverse pattern of a 3rd projection pattern is irradiated, and is a figure which shows the luminance value in the round-dot position shown to (c) (a), (b).

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
Embodiment 1 FIG.
As shown in FIG. 1, the three-dimensional measurement system includes a projector 1, an imaging device 2, a storage device 3, and a processing device 4. In the following, a case where the spatial encoding method is used as the pattern light projection method is shown, but other methods may be used.

  The projector 1 irradiates the measurement area where the measurement object is placed with the light of the plurality of encoded projection patterns and the inverted pattern thereof held under the control of the processing device 4. An example of the projector 1 is a projector. Moreover, as a projection pattern, the binary code pattern shown, for example in FIG. 5 is mentioned.

The imaging device 2 captures a measurement region in a state where light is emitted from the projector 1 according to control by the processing device 4.
The storage device 3 stores an image captured by the imaging device 2. The storage device 3 includes an HDD, a DVD, a memory, and the like.

  The processing device 4 controls the projector 1 and the imaging device 2 to perform light irradiation and imaging, and performs three-dimensional measurement of the measurement object based on the image stored in the storage device 3. At this time, the processing device 4 binarizes light and dark for each pixel based on the image, encodes a space based on the content of the binarized pattern, and performs triangulation based on the encoding result, thereby measuring objects. Calculate the three-dimensional shape of an object. The processing device 4 is executed by program processing using a CPU based on software.

Next, the configuration of a noise detection apparatus that is used in the three-dimensional measurement system and detects noise in the three-dimensional measurement will be described with reference to FIG. FIG. 2 shows a case where the function of the noise detection device is built in the processing device 4 of the three-dimensional measurement system.
As shown in FIG. 2, the noise detection apparatus includes an all-irradiation control unit 41, a non-irradiation control unit 42, a threshold setting unit 43, a threshold holding unit 44, a noise detection unit 45, and a noise removal unit 46. This noise detection apparatus is executed by program processing using a CPU based on software.

The total irradiation control unit 41 controls the projector 1 so as to irradiate the entire measurement region with light. Further, the non-irradiation control unit 42 controls the projector 1 so as not to irradiate the entire measurement region with light.
Further, the projector 1 holds, as projection patterns, a pattern that irradiates light to the entire measurement region (all irradiation pattern) and a pattern that does not irradiate light to the entire measurement region (non-irradiation pattern). The projector 1 controls the light irradiation using the two patterns according to the control by the total irradiation control unit 41 and the non-irradiation control unit 42.

  The threshold value setting unit 43 is based on the luminance values of both images captured by the imaging device 2 in a state where the irradiation of light by the projector 1 is controlled by the total irradiation control unit 41 and the non-irradiation control unit 42. A threshold for noise detection is set for each coordinate. At this time, the threshold value setting unit 43 sets a value within the range of the difference between the luminance values of both images obtained by the total irradiation control unit 41 and the non-irradiation control unit 42 (for example, a luminance value that is half the difference) as the threshold value. .

  The threshold value holding unit 44 holds data indicating the threshold value for each coordinate set by the threshold value setting unit 43. The threshold value holding unit 44 is configured by an HDD, a DVD, a memory, or the like.

  In the noise detection unit 45, the threshold value setting unit 43 sets the difference between the luminance values of both images captured by the imaging device 2 in a state where the projection pattern 1 and the light of the inverted pattern are irradiated by the projector 1 in the three-dimensional measurement. It is determined whether or not it is equal to or less than the threshold value of the corresponding coordinate stored in the threshold value holding unit 44, and when there are a predetermined number or more (for example, one or more) of projection patterns equal to or less than the threshold value, the coordinate is detected as noise. Is. In addition, when the threshold set in the threshold setting unit 43 is below the specified value, the noise detection unit 45 detects the corresponding coordinate as noise at that time.

  The noise removing unit 46 removes information on the coordinates where the noise is detected by the noise detecting unit 45.

  Next, the operation of the noise detection apparatus configured as described above will be described with reference to FIG. In the following, as a specific example, as shown in FIG. 7, the measurement object (concave workpiece 10) is arranged on the tray 11, the projector 1 is arranged obliquely above the workpiece 10, and the imaging device 2. Suppose that is placed directly above.

  In the operation of the noise detection apparatus, as shown in FIG. 3, first, the total irradiation control unit 41 controls the projector 1 so as to irradiate the entire measurement area (step ST1, total irradiation control step). Then, in accordance with this control, the projector 1 irradiates the measurement region with light of the entire irradiation pattern, and the imaging device 2 images the measurement region at that time. An image captured by the imaging device 2 is stored in the storage device 3.

  Next, the non-irradiation control unit 42 controls the projector 1 so as not to irradiate the entire measurement region with light (step ST2, no irradiation control step). Then, in accordance with this control, the projector 1 makes the measurement area non-irradiated with a non-irradiation pattern, and the imaging device 2 images the measurement area at that time. An image captured by the imaging device 2 is stored in the storage device 3.

In the measurement environment shown in FIG. 7, the case where light irradiation is controlled by the whole irradiation pattern and the non-irradiation pattern is shown in FIG. The brightness of the tray 11 in FIGS. 4 (a) and 4 (b) indicates the light irradiation area when the work 10 is not on the tray 11, the white area is the area irradiated with light, and the black area is This is a region that is not irradiated with light. In addition, FIG. 4C shows the luminance value at the point of interest (circle point in the figure) among the images taken by the imaging device 2 in the state shown in FIGS. 4A and 4B.
As shown in FIG. 4, when the concave workpiece 10 is irradiated with light of the entire irradiation pattern, the direct light located on the right side of the broken line in FIG. 4A is blocked by the side wall in front of the workpiece 10. However, since the reflected light from the back side wall reaches the point of interest, a high luminance value is obtained. Further, the non-irradiation pattern has a low luminance value.

Next, the threshold setting unit 43 is based on the luminance values of both images captured by the imaging device 2 in a state where the irradiation of light by the projector 1 is controlled by the total irradiation control unit 41 and the non-irradiation control unit 42. A threshold for noise detection is set for each coordinate of the image (step ST3, threshold setting step). At this time, the threshold setting unit 43 sets a threshold for each coordinate based on, for example, the following expression (1).
Threshold = (| luminance value obtained with all irradiation patterns−luminance value obtained with no irradiation patterns |) / 2 (1)

For example, it is assumed that the luminance value obtained with the entire irradiation pattern is 200 and the luminance value obtained with the non-irradiation pattern is 10 at the attention point shown in FIG. In this case, the threshold value at the attention point is (| 200−10 |) / 2 = 95 from the above equation (1).
In addition, when the threshold set in the threshold setting unit 43 is lower than a predetermined value (for example, 20), the noise detection unit 45 has a corresponding coordinate as an area where light of the irradiation pattern is shielded by unevenness or the like. Judgment is made, and the corresponding coordinates are detected as noise at that time.

  Data indicating the threshold value for each coordinate set by the threshold value setting unit 43 is held in the threshold value holding unit 44.

  Then, the three-dimensional measurement system performs light irradiation and imaging of each projection pattern and its inverted pattern. That is, the projector 1 irradiates the measurement area with light of each projection pattern and its inverted pattern according to control by the processing device 4. In addition, the imaging device 2 captures an image of the measurement region in a state where light irradiation is controlled by the projector 1 according to control by the processing device 4. And the memory | storage device 3 memorize | stores the image imaged by the imaging device 2. FIG.

  In this case, in the measurement environment shown in FIG. 7, when the space is divided into eight spaces by the three binary code patterns shown in FIG. 5 and the inverted pattern obtained by inverting the brightness, the irradiation image is as shown in FIGS. Become.

  Next, the noise detection unit 45 calculates the difference between the luminance values of the two images captured by the imaging device 2 in a state where the projection pattern 1 and the light of the inverted pattern are irradiated by the projector 1 in the three-dimensional measurement. It is determined whether or not the value is equal to or less than the threshold value of the corresponding coordinate set in step S5. If there are a predetermined number or more of projection patterns equal to or less than the threshold value, the coordinate is detected as noise (step ST5, noise detection step). In the following, it is assumed that noise is detected when there is one or more patterns whose luminance value difference is equal to or less than the threshold.

  For example, in the case of FIG. 8, the difference between the luminance value of the first projection pattern and the luminance value of the inversion pattern is large and larger than the threshold value. On the other hand, in the case of FIGS. 9 and 10, the difference between the luminance values of the second and third projection patterns and the luminance value of the inversion pattern is small and is equal to or less than the threshold value. Therefore, it is determined that the attention point shown in FIGS.

Next, the noise removing unit 46 removes information on the coordinates where the noise is detected by the noise detecting unit 45 (step ST6, noise removing step). There are a plurality of methods for removing information, such as adding flag information associated with each coordinate, and any method may be used.
Then, by performing this process on all the coordinates of the image, it is possible to remove, as noise, an area where there is a high possibility that an encoding error will occur.

  Thereafter, in the three-dimensional measurement system, a three-dimensional measurement of the measurement object is performed using an image stored in the storage device 3. That is, as in general encoding, if “the luminance value obtained by the projection pattern> the luminance value obtained by the inversion pattern”, the point of interest is that the light of the projection pattern is irradiated If it is vice versa, it is determined that the point is not irradiated, and encoding is performed according to the projection pattern used (in the case of a binary code pattern, encoding is performed using binary numbers, and the encoding method is a projection pattern) Depending on).

  As described above, according to the first embodiment, a case where the entire measurement region is irradiated and a case where the entire measurement region is not irradiated are picked up, and a threshold for noise detection is set for each coordinate from both images. In addition, since the region where the difference between the luminance values obtained from the irradiation result of the projection pattern and the reverse pattern is equal to or less than the threshold value is removed as noise, three-dimensional measurement is performed on the measurement object by the pattern light projection method. In such a case, noise due to unevenness can be detected.

  In the above, when there is one or more patterns (projection patterns, inversion patterns) whose luminance value difference is equal to or less than the threshold, the coordinates are determined to be coordinates that are highly likely to be noise. . However, the present invention is not limited to this, and it may be determined that noise is present when there are a plurality of (for example, two) or more patterns whose luminance value difference is equal to or less than a threshold value.

  In the above description, a binary code pattern is used as the projection pattern. However, the present invention is not limited to this, and the present invention can be applied to any three-dimensional measurement using the pattern light projection method. For example, a gray code pattern or the like may be used as the projection pattern.

  In the present invention, any constituent element of the embodiment can be modified or any constituent element of the embodiment can be omitted within the scope of the invention.

DESCRIPTION OF SYMBOLS 1 Projector 2 Imaging device 3 Memory | storage device 4 Processing apparatus 10 Recessed work 11 Tray 41 Total irradiation control part 42 Non-irradiation control part 43 Threshold setting part 44 Threshold holding part 45 Noise detection part 46 Noise removal part

Claims (4)

A projector that irradiates light of a plurality of projection patterns and its inversion pattern to a measurement region on which a measurement object is placed, an imaging device that images the measurement region in a state where light is irradiated by the projector, In a noise detection apparatus that is used in a three-dimensional measurement system including a processing device that performs three-dimensional measurement of the measurement object based on an image captured by the imaging device, and detects noise in the three-dimensional measurement,
A total irradiation control unit for controlling the projector to irradiate the entire measurement region with light;
A non-irradiation control unit for controlling the projector so as not to irradiate light on the entire measurement region;
Noise is applied to each coordinate of the image based on the luminance values of both images captured by the imaging device in a state in which light irradiation by the projector is controlled by the total irradiation control unit and the non-irradiation control unit. A threshold setting unit for setting a threshold for detection;
In the three-dimensional measurement, a difference between luminance values of both images captured by the imaging device in a state where the light of the projection pattern and its inverted pattern is irradiated by the projector is a corresponding coordinate set by the threshold setting unit. A noise detection unit that determines whether or not the coordinates are equal to or less than a predetermined number of projection patterns that are equal to or less than the threshold,
A noise detection apparatus comprising: a noise removal unit that removes information on coordinates where noise is detected by the noise detection unit. The threshold value setting unit sets, as the threshold value, a value within a range of a difference between luminance values of both images obtained by the control by the total irradiation control unit and the non-irradiation control unit. Noise detector. The noise detection device according to claim 1, wherein the noise detection unit detects the corresponding coordinate as noise when the threshold set by the threshold setting unit is below a predetermined value. A projector that irradiates light of a plurality of projection patterns and its inversion pattern to a measurement region on which a measurement object is placed, an imaging device that images the measurement region in a state where light is irradiated by the projector, In a noise detection method for detecting noise in the three-dimensional measurement, used in a three-dimensional measurement system including a processing device that performs three-dimensional measurement of the measurement object based on an image captured by the imaging device.
A total irradiation control unit for controlling the projector to irradiate the entire measurement area with light;
A non-irradiation control unit controls the projector so as not to irradiate light to the entire measurement region, and
Based on the brightness values of both images captured by the imaging device in a state where the irradiation of light by the projector is controlled by the total irradiation control unit and the non-irradiation control unit, A threshold setting step for setting a threshold for noise detection with respect to coordinates;
In the three-dimensional measurement, the noise detection unit is configured such that a difference between luminance values of both images captured by the imaging device in a state where the projection pattern and its inverted pattern are irradiated by the projector is determined by the threshold setting unit. A noise detection step of determining whether or not the set coordinate is less than or equal to the threshold value, and detecting the coordinate as noise when the projection pattern that is less than or equal to the threshold value is greater than or equal to the specified number;
A noise removal method, wherein the noise removal unit includes a noise removal step of removing information on coordinates where noise is detected by the noise detection unit.

Images