Three-dimensional measuring device

Abstract

This three-dimensional measuring device comprises a light projecting unit for projecting reference light onto a target object while scanning the reference light at a predetermined angular velocity, a light receiving unit for receiving the reference light reflected from the target object, and a three-dimensional information calculating unit for calculating three-dimensional information of the target object by means of triangulation based on information of the light receiving unit, wherein the reference light is diffused light having a predetermined diffusion angle.

Description

[Technical Field]

[0001] The present invention relates to a three-dimensional measurement technique, and more particularly to an optical scanning three-dimensional measurement device. [Background technology]

[0002] Many of the 3D measurement devices commercialized for industrial use are based on triangulation using structured illumination. Various light projection methods have been proposed, such as the phase shift method and spatial code method, and one type of these is the light section method, which scans with a slit of light. The light section method is generally said to have good distance measurement accuracy, but compared to the phase shift method, spatial code method, etc., it requires a large number of images to be captured in one measurement, which means that the measurement takes a long time.

[0003] Meanwhile, event-based image sensors, which are based on a different concept from general image sensors (hereafter referred to as frame-based image sensors), have been proposed. While frame-based image sensors capture and output images by exposing for a fixed period of time, event-based image sensors monitor changes in illuminance from moment to moment for each pixel independently. When they detect a change in illuminance above a preset threshold, they output event information, including the pixel's position, the time of the change, and the polarity of the change (whether it became brighter or darker). Because event-based image sensors only output information from pixels where there has been a change in illuminance, they are faster than frame-based image sensors, which output information from all pixels. Therefore, the use of event-based image sensors in three-dimensional measurement devices using the light-section method has been proposed.

[0004] In the light-section method using an event-based image sensor, the time at which a change in illuminance occurs at each pixel is observed, but even with a slit light scanned at the same angular velocity, the change in illuminance observed on the light-receiving surface of the image sensor becomes faster the farther the subject is, making it more difficult to observe a change in illuminance on the far side than a change in illuminance on the near side. In other words, the speed of change in illuminance on the far side is a constraint on the design of the three-dimensional measurement device. The following documents, for example, are known as background art related to this application.

[0005] Patent document 1 describes a method for appropriately performing three-dimensional measurements of an object to be measured regardless of the magnitude and diversity of the surface reflectance of the object to be measured, in which the irradiation intensity of slit light is changed in stages, the three-dimensional shape of the object to be measured for each irradiation intensity is measured using two-dimensional images containing light-section lines for each irradiation intensity, and the optimal three-dimensional shape of the object to be measured is determined based on the three-dimensional shape results for each irradiation intensity.

[0006] Patent Document 2 describes that an event camera is used in a phase shift method that uses brightness values, and brightness information is calculated based on the time difference between the output of event data for a positive brightness change and the output of event data for a negative brightness change on a pixel-by-pixel basis in the captured image.

[0007] Patent document 3 describes a method of measuring the shape of an object to be measured multiple times by scanning a slit laser beam so that at least a portion of each measurement area overlaps, and integrating the measurement data using reliability information between the multiple measurement data obtained.

[0008] Patent Document 4 discloses a three-dimensional input device that captures images synchronously using two image sensors while scanning an object with a slit light, stores the frame number at which the maximum brightness is detected for each pixel of each image sensor, matches pixels based on the frame numbers, and calculates distance based on the matched pixels. [Prior art documents] [Patent documents]

[0009] [Patent Document 1] Japanese Patent Application Laid-Open No. 2014-181995 [Patent Document 2] Patent Publication No. 2021-067644 [Patent Document 3] Japanese Patent Application Laid-Open No. 2008-002995 [Patent Document 4] Japanese Patent Application Laid-Open No. 2000-088539 Summary of the Invention [Problem to be solved by the invention]

[0010] In view of the problems of the prior art, the present invention aims to provide a technique for optical scanning three-dimensional measurement that enables three-dimensional measurement under generally constant conditions regardless of the distance to the object. [Means for solving the problem]

[0011] One aspect of the present disclosure is a scanning optical system including a light projecting unit that projects a reference light onto an object while scanning the object with the reference light at a predetermined angular velocity, a light receiving unit that receives the reference light reflected from the object, and a three-dimensional information calculating unit that calculates three-dimensional information of the object by triangulation based on information from the light receiving unit, wherein the reference light is diffused light having a predetermined diffusion angle. the light receiving unit has a plurality of pixels, detects a change in illuminance when illuminating an object with reference light for each pixel, and outputs the position of the pixel where the illuminance change occurred, the time of the change, and the polarity of the change as event information; A three-dimensional measuring device is provided. [Effects of the Invention]

[0012] According to one aspect of the present disclosure, the reference light is diffused light having a predetermined diffusion angle, so that three-dimensional measurement can be performed under substantially constant conditions regardless of the distance to the object. [Brief explanation of the drawings]

[0013] [Figure 1A] FIG. 1 is an explanatory diagram illustrating the measurement principle of the stereo method. [Figure 1B] FIG. 1 is an explanatory diagram illustrating the measurement principle of the stereo method. [Figure 2A] FIG. 1 is a schematic perspective view of a three-dimensional measuring device using a light-section method. [Figure 2B] FIG. 1 is an explanatory diagram illustrating the measurement principle of the light-section method. [Figure 2C] 10A and 10B are diagrams showing examples of reference light images captured while changing the projection angle; [Figure 3] FIG. 1 is an explanatory diagram showing the measurement principle of the light-section method using an event-based image sensor. [Figure 4A] FIG. 1 is an explanatory diagram illustrating the difference between a normal light-section method and a light-section method using an event-based image sensor. [Figure 4B] FIG. 1 is an explanatory diagram illustrating the difference between a normal light-section method and a light-section method using an event-based image sensor. [Figure 5] 1 is a configuration diagram of a three-dimensional measurement apparatus according to an embodiment; [Figure 6] 10 is a graph showing an example of the illuminance distribution of the reference light. [Figure 7A] FIG. 2 is a side view showing the center line of a reference beam projected onto an object. [Figure 7B] FIG. 10 is an image diagram showing the center line of a reference beam projected onto an object. [Figure 8A] FIG. 10 is a side view showing the center line of a reference beam projected onto a corner of an object. [Figure 8B] FIG. 10 is an image diagram showing the center line of a reference beam projected onto a corner of an object. [Figure 9] 10 is a graph showing a change in illuminance of a reference light in one pixel. [Figure 10] 1 is a graph showing the occurrence of events in one pixel. [Figure 11A] FIG. 10 is an explanatory diagram illustrating the relationship between the thickness of the slit light and the field of view range. [Figure 11B] FIG. 10 is an explanatory diagram illustrating the relationship between the thickness of the reference light and the field of view range. [Figure 12A] FIG. 2 is a perspective view of a light projecting unit according to an embodiment of the present invention; [Figure 12B] FIG. 2 is a side view of the light projecting unit according to the embodiment. [Figure 12C] FIG. 2 is a top view of a light projecting unit according to an embodiment. [Figure 13A] FIG. 10 is a perspective view of a light projecting unit according to a modified example. [Figure 13B] FIG. 10 is a side view of a light projecting unit according to a modified example. [Figure 13C] FIG. 10 is a top view of a light projecting unit according to a modified example. [Figure 14A] FIG. 10 is a top view of a light projecting unit according to a further modified example. [Figure 14B] FIG. 10 is a top view of a light projecting unit according to a further modified example. [Figure 14C] FIG. 10 is a top view of a light projecting unit according to a further modified example. [Figure 15] FIG. 1 is a block diagram of a three-dimensional measurement apparatus according to an embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0014] Hereinafter, embodiments of the present disclosure will be described in detail with reference to the accompanying drawings. In each drawing, the same or similar components are assigned the same or similar reference numerals. Furthermore, the embodiments described below do not limit the technical scope and meaning of the terms of the invention described in the claims.

[0015] To facilitate understanding of the three-dimensional measurement device of this embodiment, the measurement principles of the stereo method, the light-section method, and the light-section method using an event-based image sensor will first be described in order. FIGS. 1A and 1B are explanatory diagrams illustrating the measurement principle of the stereo method. In the stereo method, the left and right cameras are arranged, for example, in an isotropic parallel arrangement. That is, the two cameras are separated by a base length B, the left and right optical axes of both cameras are arranged parallel, and the left and right imaging surfaces of the same size are arranged in a plane perpendicular to the optical axes, with the x and y directions of each imaging surface oriented in the same direction. Alternatively, the imaging position and imaging orientation of a single camera may be changed to such an isotropic parallel arrangement, and images may be captured sequentially. The left and right imaging surfaces are, for example, configured using an image sensor in which multiple pixels are two-dimensionally arranged, but may also be configured using an image sensor in which multiple pixels are one-dimensionally arranged (for example, arranged only in the x direction).

[0016] The image of point P on an object in the target space is formed at the intersection of the line connecting the focal points of the left and right cameras to point P and the imaging plane of each camera. When an image is captured using only one of the left and right cameras, it is clear that point P is on the line connecting the focal points to the image, but the distance to point P is unknown. However, when an image is captured using both the left and right cameras, it is clear that point P is on the intersection of the two lines connecting the focal points to the image, and the distance to point P can be calculated from the similarity relationship between the triangle enclosed by the two lines and the baseline connecting the focal points of the left and right cameras, and the triangle enclosed by the two lines and the images of point P captured by the left and right cameras (this is known as triangulation). Here, the distance to point P refers to the depth from the baseline connecting the focal points of both cameras to point P in the direction of the optical axes of the left and right cameras. In other words, the distance to point P is the Z coordinate of point P in a camera coordinate system fixed to the focal position of one of the left and right cameras.

[0017] If the image coordinates of the image of point P on the left camera are ml(xl, yl) and the image coordinates of the image of point P on the right camera are mr(xr, yr), then the difference in the x direction between the images of point P on the left and right cameras (the so-called parallax) is D = xl - xr. If the baseline length between the left and right cameras is B and the focal length of both cameras is f, the distance Z to point P (the Z coordinate of point P in the camera coordinate system) can be calculated using the similarity relationship of triangles using the following formula. In the formula, the baseline length B and focal length f are constants determined by the camera design, so once the parallax D is found, the distance to point P (the Z coordinate of point P in the camera coordinate system) can be calculated.

[0018]

number

[0019] Similarly, from the similarity relationship of triangles, the X coordinate (X = (xr - cx) Z / f) and Y coordinate (Y = (yr - cy) Z / f) of point P in the camera coordinate system can also be calculated. In the formula, cx and cy are the image coordinates of the optical axis center Or of the right camera.

[0020] Next, the measurement principle of the light-section method will be explained. Figure 2A is a schematic perspective view of a three-dimensional measurement device using the light-section method. A three-dimensional measurement device 1 using the light-section method replaces the left camera in Figures 1A and 1B with a light-projection unit 2. In the light-section method, the light-projection unit 2 projects reference light L onto an object W in a target space S while scanning, and the light-receiving unit 3 receives the reference light L reflected by the object W. Instead of calculating the parallax D, the three-dimensional position of any point on the object W (and ultimately the three-dimensional shape of the object W) is calculated based on the projection angle of the reference light L. The light-section method is said to have good accuracy because it does not require the detection of corresponding points P on the left and right images as is necessary in the stereo method for calculating the parallax D.

[0021] The light projection unit 2 includes a light source 20 of the reference light L, projection optical systems 21-22 that shape the beam of the reference light L, and a scanning unit 23 that scans the reference light L. The light source 20 is configured, for example, by a semiconductor laser, but is not limited to this. In other embodiments, the light source 20 may be configured by other light sources such as a solid-state laser (fiber laser, YAG laser, etc.) or a gas laser (carbon dioxide laser, helium-neon laser, argon laser, etc.). The projection optical systems 21-22 are configured, for example, by beam-shaping lenses such as collimating lenses and cylindrical lenses. The scanning unit 23 is configured, for example, by a scanner including a galvanometer scanner, an encoder, etc. The reference light L is emitted from the light source 20, shaped into a slit beam by the projection optical systems 21-22, scanned by the scanning unit 23, and projected onto the object W.

[0022] The light receiving unit 3 includes a light receiving optical system 30 that receives light reflected from the object W, and an image sensor 31. The light receiving optical system 30 includes, for example, a condenser lens. The image sensor 31 includes, for example, a plurality of pixels arranged two-dimensionally, but is not limited to this, and in other embodiments, the image sensor 31 may be composed of a plurality of pixels arranged one-dimensionally. The reference light L reflected from the object W is collected by the light receiving optical system 30 and received by the image sensor 31.

[0023] Figure 2B is an explanatory diagram showing the measurement principle of the light-section method. The center of rotation of the reference light L is, although not limited to, assumed to be at the left focal point of the left camera in the stereo method. If the projection angle of the reference light from the left optical axis is θ (clockwise is positive), the position xl of the image of point P on the virtual plane can be found from the following equation:

[0024]

number

[0025] Substituting Equation 2 into Equation 1 gives the following equation:

[0026]

number

[0027] 2C is a diagram showing an example of reference light images I1 to I4 captured while changing the projection angle θ. During distance measurement, multiple images are captured while changing the projection angle θ of the reference light L, and the position xr of the image S of the reference light L on the object W is detected for each row of each of the images I1 to I4. The projection angle θ when the image was captured and the detected position xr of the image S of the reference light L are substituted into Equation 3 to calculate the distance Z to point P (the Z coordinate of point P in the camera coordinate system). The value of the projection angle θ is obtained from an encoder or the like of the scanning unit 23 (see FIG. 2A).

[0028] That is, instead of calculating the parallax D=xl-xr as in the stereo method, the light-section method calculates the distance to point P on the object W (the Z coordinate of point P in the camera coordinate system) based on the projection angle θ of the reference light L at each time and the position xr of each pixel through which the reference light L passed. Similarly, the X coordinate (X=(xr-cx)·Z / f) and Y coordinate (Y=(yr-cy)·Z / f) of point P in the camera coordinate system are also calculated from the similarity relationship of the triangle. In the formula, cx and cy are the image coordinates of the center of the optical axis of the light-receiving unit 3.

[0029] Next, the measurement principle of the light-section method using an event-based image sensor will be explained. Figure 3 is an explanatory diagram showing the measurement principle of the light-section method using an event-based image sensor. In the light-section method using an event-based image sensor, the light-receiving unit 3 of the light-section method (see Figure 2A) is replaced with an event-based image sensor. If the reference light L is rotated at a uniform speed of a predetermined angular velocity ω and the elapsed time from the moment the reference light L passes through the left optical axis is taken as t, the projection angle θ of the reference light L at time t can be calculated using the following equation.

[0030]

number

[0031] Substituting Equation 4 into Equation 3 gives the following equation:

[0032]

number

[0033] When the reference light L is scanned during distance measurement, an event occurs at each pixel of the event-based image sensor due to changes in illuminance at each pixel. The event-based image sensor outputs the position of the pixel where the illuminance change occurred, the time of the change, and the polarity of the change as event information. The time t at which the reference light L passed through each pixel is estimated from the event information, and the position xr of each pixel where the event occurred and the time t at which the reference light L passed through that pixel are substituted into equation 5 to calculate the distance Z to point P (the Z coordinate of point P in the camera coordinate system). Similarly, the X coordinate of point P in the camera coordinate system (X = (xr - cx) Z / f) and the Y coordinate of point P (Y = (yr - cy) Z / f) are also calculated from the similarity of triangles. In the equation, cx and cy are the image coordinates of the center of the optical axis of the light receiving unit 3.

[0034] In other words, the light-section method using an event-based image sensor outputs only event information when a change in illuminance occurs in each pixel, instead of outputting multiple images for each projection angle θ as in the ordinary light-section method using a frame-based image sensor, and is therefore characterized by the ability to perform three-dimensional measurement at high speed.

[0035] 4A and 4B are explanatory diagrams illustrating the difference between a normal light-section method using a frame-based image sensor and a light-section method using an event-based image sensor. In summary, with the normal light-section method using a frame-based image sensor, distance measurement is performed by detecting positions x1 and x2 of the images of the reference light L at points P1 and P2 on the object W for each projection angle θ of the reference light L, as shown in FIG. 4A. On the other hand, with the light-section method using an event-based image sensor, distance measurement is performed by detecting times t1 and t2 at which the reference light L passes through points P1 and P2 on the object W for each pixel, as shown in FIG. 4B. It should be noted that the three-dimensional measuring device 1 of this embodiment can be used for either measurement principle: the light-section method using a frame-based image sensor, which captures images at a constant cycle and determines the time of passage of the reference light L from the change in brightness of each pixel over time to measure distance; or the light-section method using an event-based image sensor.

[0036] The three-dimensional measuring device 1 of this embodiment will be described below. FIG. 5 is a configuration diagram of the three-dimensional measuring device 1 of this embodiment. The three-dimensional measuring device 1 is also called a three-dimensional camera or three-dimensional sensor, and is used, for example, as a visual sensor for machines such as robots and vehicles. A reference light L is projected onto an object W located within a viewing angle φ and a distance measurement range R of a light-receiving unit 3 while being scanned by a light-projecting unit 2 at a predetermined angular velocity ω. The projected reference light L is reflected by the object W and received by the light-receiving unit 3. The three-dimensional measuring device 1 calculates three-dimensional information of the object W based on the principle of triangulation, as described above. Note that the term "three-dimensional information" used herein is not limited to the distance to an arbitrary point on the object W or the three-dimensional coordinates, but also includes point cloud data and shape data of the object W.

[0037] Furthermore, as will be described in detail below, the reference light L used in the three-dimensional measurement apparatus 1 of this embodiment is a slit light having a predetermined diffusion angle in the thickness direction as well, but it should be noted that this is not limited to this and that in other embodiments, the reference light L may be other forms of reference light, such as a spot light having a predetermined diffusion angle. The reference light L has an illuminance distribution that is maximum near its center.

[0038] Fig. 6 is a graph showing an example of the illuminance distribution of reference light L in the thickness direction, and Figs. 7A and 7B are a side view and an image diagram showing the center line C of reference light L projected onto object W. Please note that in Figs. 7A and 7B, the reference light L is depicted as an extremely thick bright line for ease of understanding. When reference light L having the illuminance distribution shown in Fig. 6 is irradiated onto object W as shown in Fig. 7A and an image is acquired, the reference light L appears on the image as a bright line with a certain width, as shown in Fig. 7B.

[0039] In the normal light-section method using a general frame-based image sensor, an image of the illuminance distribution as shown in Fig. 7B is output from the frame-based image sensor, and the position of the image of the reference light L on the light-receiving surface is detected from that image by detecting the center line C of the image of the reference light L. Generally, a weighted average of the illuminance in the width direction of the image of the reference light L is used as a method for detecting the center line C of the image of the reference light L.

[0040] Generally, a collimated slit beam with a roughly uniform thickness (light with a thin beam that does not spread) is used as the reference beam L in a conventional light-section method. There is a reason for this. Figures 8A and 8B are a side view and an image diagram showing the center line C of the reference beam L projected onto a corner of the object W. When the entire width of the reference beam L strikes the object W as shown in Figure 7A, the entire width of the reference beam L is captured in the image as shown in Figure 7B. Therefore, the center line C can be accurately detected by taking a weighted average of the illuminance in the width direction of the image of the reference beam L. On the other hand, when the reference beam L strikes the corner of the object W as shown in Figure 8A, only the portion that strikes the top surface of the object W is captured in the image as shown in Figure 8B. Therefore, the center line C of the detected image of the reference beam L is shifted from the center of the actual reference beam L, resulting in a corresponding distance measurement error. In other words, to measure the distance as accurately as possible around uneven areas such as steps, it is desirable for the reference beam L to be a collimated slit beam with as thin a thickness as possible.

[0041] However, this common technical knowledge regarding the light-section method is a requirement necessary when attempting to detect the center line C of the image of reference light L using a spatially weighted average of illuminance, etc., but does not apply when attempting to detect the center line C of the image of reference light L using a temporally weighted average of illuminance at individual pixels, or when attempting to detect the center line C of the image of reference light L using temporal changes in illuminance, as in the light-section method using an event-based image sensor. The reason for this will be explained in detail below.

[0042] In the case of the light-section method using an event-based image sensor, a similar image of the reference light L is formed on the light-receiving surface, but since there is no change in illuminance when the reference light L is stationary, no event is output from the event-based image sensor and no information can be obtained. By scanning the reference light L, a change in illuminance occurs at the pixel where the image of the reference light L moves, and an event is output.

[0043] Figure 9 is a graph showing the change in illuminance of the reference light L at one pixel, and Figure 10 is a graph showing the occurrence of events at one pixel. Focusing on one pixel of an event-based image sensor, when the reference light L passes through a specific part on the object W captured in the pixel, a change in illuminance i is observed, where illuminance i increases as the center of the reference light L approaches that part and decreases as it moves away, as shown in Figure 9. The event-based image sensor generates an event when it detects an increase or decrease in illuminance above a preset threshold. For example, as shown in Figure 10, two positive events occur in conjunction with the increase in illuminance as the reference light L passes, and one negative event occurs in conjunction with the decrease in illuminance as the reference light L passes and moves away. Then, in the light-section method using an event-based image sensor, instead of detecting the position of the center line C of the image of reference light L at a certain time, the time when the center line C of the image of reference light L passed through each pixel is estimated based on the occurrence times of these events (e.g., t1, t2, t3) and the polarity of the change in illuminance (e.g., positive polarity, positive polarity, and negative polarity). In other words, in the light-section method using an event-based image sensor, the time when the center line C of the image of reference light L passed through that pixel is estimated based on the rise and fall of illuminance.

[0044] Therefore, in the light-section method using an event-based image sensor, the time at which the reference light passes is estimated based on the change in illuminance at each pixel over time, and is therefore independent of the relationship with the illuminance of surrounding pixels. In other words, even if the image of the reference light appears thick on the light-receiving surface of the image sensor, the problems described with reference to Figures 8A and 8B do not occur. In other words, the reference light L does not need to be a thin, collimated slit light.

[0045] On the other hand, with the light-section method using an event-based image sensor, the speed of change in illuminance must be taken into consideration because the change in illuminance over time is observed. The speed of change in illuminance observed by a pixel depends on the width and movement speed of the image of reference light L on the light-receiving surface. For example, if the width of the image of reference light L on the light-receiving surface is w and the movement speed of the image of reference light L is v, the time T required for the image of reference light L to pass a certain pixel can be expressed by the following equation:

[0046]

number

[0047] In other words, the time during which a change in illuminance is observed at a certain pixel is only during the required transit time T, and the image sensor is required to have high speed to capture the change in illuminance that occurs within the short period of time required for transit T. Here, we consider the width w and moving speed v of the image of the reference light L on the light-receiving surface, which affect the required transit time T.

[0048] Figure 11A is an explanatory diagram illustrating the relationship between the thickness of the slit light beam L and the field of view. First, we consider the width w of the image of the slit light beam L, assuming the use of collimated light, which has a roughly constant thickness, in accordance with the accepted theory of light-section imaging. Because collimated light beams do not spread, their thickness remains roughly constant regardless of the distance to the object W. (However, even collimated light generally has its own divergence angle. However, since the divergence angle can be very small depending on the type of light source, we will ignore the divergence angle in this document.) Meanwhile, the light-receiving unit 3 has a field of view φ, so the field of view increases with distance. In other words, when the distance to the object W increases n times, the field of view of the light-receiving unit 3 increases n times vertically and horizontally. However, because the thickness of the reference light beam L is constant, the relative thickness of the reference light beam L relative to the field of view of the light-receiving unit 3 decreases by 1 / n times when the distance to the object W increases by n times. Therefore, the width w of the image of the reference light beam L captured on the image also decreases by 1 / n times. That is, the farther the object W is, the thinner the reference light L appears in the image.

[0049] Applying this to Equation 6 above, the width w of the image of the reference light L becomes 1 / n times larger, and the required transit time T required for the image of the reference light L to pass through a certain pixel also becomes 1 / n times larger. In other words, the longer the distance to the object W, the shorter the required transit time T, and therefore the image sensor 31 is required to be correspondingly faster. In other words, in the design of the three-dimensional measuring apparatus 1, even if there is room on the closer side, the farther side becomes a design constraint. To relax this constraint and perform three-dimensional measurement under roughly constant conditions regardless of the distance to the object W, the three-dimensional measuring apparatus 1 of this embodiment imparts a predetermined diffusion angle to the reference light L.

[0050] 11B is an explanatory diagram illustrating the relationship between the thickness of the reference light L and the field of view range. By providing the reference light L with a divergence angle Ψ, the thickness of the reference light L increases in proportion to the distance to the object W. This makes it possible to maintain the relative thickness of the reference light L with respect to the field of view range of the light receiving unit 3, and therefore the width w of the image of the reference light L on the image sensor 31 (light receiving surface), approximately constant, regardless of the distance to the object W.

[0051] Meanwhile, considering the moving speed v of the image of the reference light L on the image sensor 31 (light receiving surface), the reference light L is rotated at a constant speed by the light projecting unit 2 at a predetermined angular velocity ω. If the distance from the center of rotation of the reference light L to the object W is r, the moving speed u of the reference light L on the object W is u = rω. If the distance from the focal point of the light receiving unit 3 to a certain point on the object W is Z and the projection angle of the reference light with respect to the optical axis of the light receiving unit 3 at that time is θ, then r = Z / cosθ, and therefore u = Zω / cosθ. Furthermore, if the focal length of the light receiving unit 3 is f, then the moving speed v of the image of the reference light L on the light receiving surface is v = u × f / Z = fω / cosθ. Therefore, the moving speed v of the image of the reference light L on the light receiving surface does not depend on the distance Z to a certain point on the object W.

[0052] Applying this to the above equation 6, by giving the reference light L a predetermined diffusion angle Ψ, the width w of the image of the reference light L on the light-receiving surface does not depend on the distance to the object W, and the movement speed v of the image of the reference light L on the light-receiving surface does not substantially depend on the distance to the object W, so the time T required for the image of the reference light L to pass through each pixel on the light-receiving surface can be kept approximately constant.

[0053] Since the width w of the image of the reference light L on the light-receiving surface is proportional to the magnitude of the divergence angle Ψ of the reference light L, the time T required for the image of the reference light L to pass a certain pixel on the light-receiving surface is proportional to the divergence angle Ψ of the reference light L. Furthermore, since the moving speed v of the image of the reference light L on the light-receiving surface is proportional to the angular velocity ω of the reference light L, the time T required for the image of the reference light L to pass a certain pixel on the light-receiving surface is inversely proportional to the angular velocity ω of the reference light L.

[0054] The moving speed v (= fω / cosθ) of the image of the reference light L on the light-receiving surface also depends on the projection angle θ of the reference light L at any given time. The range of the projection angle θ depends on the layout of the three-dimensional measurement device 1. For example, if the projection angle θ required to scan the entire field of view of the light-receiving unit 3 is designed to range from -10 degrees to 20 degrees, the range of cosθ will be 0.94 to 1.0. This is when looking at the entire field of view of the light-receiving unit 3, but when looking at each pixel, the range of θ over which the reflected light of the reference light L reflected by the object W forms an image on that pixel is further narrowed. Therefore, differences in the projection angle θ of the reference light L do not significantly affect the transit time T required for the image of the reference light L on the light-receiving surface to pass through each pixel.

[0055] In the context of the three-dimensional measurement device 1, there is a desire to minimize the time required for imaging, i.e., the imaging time required to scan the entire field of view φ with the reference light L. If the range of projection angle θ of the reference light L required to scan the entire field of view φ is Δθ, then the imaging time τ is τ = Δθ / ω. Since the projection angle range Δθ is determined by the layout of the three-dimensional measurement device 1 (in the previous example, 30 degrees, from -10 to 20 degrees), shortening the imaging time τ requires increasing the angular velocity ω. As shown in Equation 6 above, the transit time T is calculated as T = w / v. Increasing the angular velocity ω of the reference light L proportionally increases the moving velocity v of the image of the reference light L on the light-receiving surface, thereby shortening the transit time T. However, since the image sensor 31 is an electronic circuit, its speed is limited. Therefore, it is necessary to ensure that the transit time T required for the image of the reference light L to pass a certain pixel on the light-receiving surface does not exceed the limits of the image sensor 31.

[0056] Therefore, it is advisable to increase the width W of the reference light L by an amount corresponding to an increase in the moving speed v of the image of the reference light L on the light-receiving surface. In other words, by increasing the divergence angle Ψ of the reference light L in the thickness direction by an amount corresponding to an increase in the angular velocity ω, the imaging time τ can be shortened without changing the transit time T required for the image of the reference light L to pass a certain pixel on the light-receiving surface. For example, when designing the three-dimensional measurement device 1, it is advisable to first set a design lower limit Tmin of the transit time T based on the high speed of the image sensor 31, and then determine the divergence angle Ψ and angular velocity ω of the reference light L based on the design lower limit Tmin of the transit time. However, this is effective only when the width w of the image of the reference light L projected on the light-receiving surface is designed to be constant regardless of the distance to the object W by providing the divergence angle Ψ in the thickness direction of the reference light L.

[0057] The following describes the configuration of the light projector 2, which provides a predetermined divergence angle Ψ in the thickness direction of the reference light L. Figures 12A-12C are perspective, side, and top views of one embodiment of the light projector 2. In this example, a semiconductor laser is used as the light source 20. Generally, semiconductor lasers emit light from their edge surfaces, resulting in astigmatic differences (astigmatism), which result in different divergence angles in directions perpendicular and parallel to the edge surface, resulting in an elliptical beam. To facilitate beam shaping, the reference light L emitted from the light source 20 is first collimated by a collimating lens 21, and astigmatism is corrected. In other words, the collimating lens 21 is preferably a convex aspherical lens that corrects astigmatism. The collimating lens 21 has an incident surface 21a and an exit surface 21b for the reference light L, with the incident surface 21a being a convex aspherical surface and the exit surface 21b being a flat surface. However, the collimator lens 21 is not limited to this, and it is sufficient that at least one of the entrance surface 21 a and the exit surface 21 b is configured as a convex aspherical surface, thereby allowing the collimator lens 21 to temporarily shape the reference light L into collimated light without astigmatism.

[0058] The reference light L shaped by the collimator lens 21 is shaped by the beam shaping lens 22 into a slit beam having a predetermined divergence angle Ψ in the thickness direction (X direction) of the reference light L. In other words, the beam shaping lens 22 is a concave-convex lens that shapes the reference light L into a slit beam having a predetermined divergence angle Ψ. The beam shaping lens 22 has an incident surface 22a and an exit surface 22b for the reference light L, and the incident surface 22a is configured as a concave-convex surface that is convex along the X direction perpendicular to the optical axis of the reference light L and concave along the Y direction perpendicular to the optical axis of the reference light L and the X direction, and the exit surface 22b is configured as a concave surface that is concave along the Y direction. However, the beam shaping lens 22 is not limited to this, and it is sufficient that at least one of the incident surface 22a and the exit surface 22b is configured as a convex surface that is convex along the X direction and at least one of the incident surface 22a and the exit surface 22b is configured as a concave surface that is concave along the Y direction. As a result, the beam shaping lens 22 converges the reference light L in the X direction to give it a predetermined diffusion angle Ψ, and diffuses the reference light L in the Y direction to form a slit beam. Alternatively, in another embodiment, the beam shaping lens 22 may be configured such that at least one of the entrance surface 22a and the exit surface 22b is a concave surface that is concave along the X direction and the Y direction. As a result, the beam shaping lens 22 diffuses the reference light L in the X direction to give it a predetermined diffusion angle Ψ, and diffuses the reference light L in the Y direction to form a slit beam. The magnitude of the diffusion angle Ψ is determined by the radius of curvature of the entrance surface 22a or the exit surface 22b.

[0059] The reference light L, which has been shaped by the beam shaping lens 22 into a slit light having a predetermined divergence angle Ψ in the X direction (thickness direction of the reference light L), is projected onto the target W while being scanned by a scanning unit 23 (see FIG. 2A) equipped with a galvanometer scanner or the like. In the example of FIGS. 12A to 12C, a combination of multiple lenses forms the slit light having a predetermined divergence angle Ψ in the thickness direction of the reference light L, but this is not limiting, and a single lens may be used to form the slit light having a predetermined divergence angle Ψ in the thickness direction of the reference light L.

[0060] Below, we will explain a modified example in which a single lens is used to form a slit beam having a predetermined divergence angle Ψ in the thickness direction of the reference beam L. FIGS. 13A-13C are perspective, side, and top views of the light projector 2 of the modified example. In this modified example, a solid-state laser (fiber laser, YAG laser, etc.) or a gas laser (carbon dioxide laser, helium-neon laser, argon laser, etc.) is assumed as the light source 20. Lasers such as solid-state lasers and gas lasers generally have a very small divergence angle and are often emitted as collimated light. Therefore, in this modified example, the collimating lens 21 is not necessary. Furthermore, since lasers such as solid-state lasers and gas lasers generally have a relatively large output compared to semiconductor lasers, changes in illuminance can be observed well even on a distant object W.

[0061] The reference light L emitted from the light source 20 is shaped by the beam shaping lens 22 into a slit beam having a predetermined divergence angle Ψ in the thickness direction (X direction) of the reference light L. The beam shaping lens 22 is configured with substantially the same shape as that described with reference to FIGS. 12A to 12C. That is, the beam shaping lens 22 has an incident surface 22a and an exit surface 22b for the reference light L, and the incident surface 22a is configured with an uneven surface that is convex along the X direction perpendicular to the optical axis of the reference light L and concave along the Y direction perpendicular to the optical axis of the reference light L and the X direction, and the exit surface 22b is configured with a concave surface that is concave along the Y direction. However, the beam shaping lens 22 is not limited to this, and it is sufficient that at least one of the incident surface 22a and the exit surface 22b is configured with a convex surface that is convex along the X direction and at least one of the incident surface 22a and the exit surface 22b is configured with a concave surface that is concave along the Y direction. As a result, the beam shaping lens 22 converges the reference light L in the X direction to give it a predetermined diffusion angle Ψ, and diffuses the reference light L in the Y direction to form a slit beam. Alternatively, in another embodiment, the beam shaping lens 22 may be configured such that at least one of the entrance surface 22a and the exit surface 22b is a concave surface that is concave along the X direction and the Y direction. As a result, the beam shaping lens 22 diffuses the reference light L in the X direction to give it a predetermined diffusion angle Ψ, and diffuses the reference light L in the Y direction to form a slit beam. The magnitude of the diffusion angle Ψ is determined by the radius of curvature of the entrance surface 22a or the exit surface 22b.

[0062] The embodiments of FIGS. 12A-12C and the modified examples of FIGS. 13A-13C are examples in which a slit beam having a predetermined divergence angle Ψ is formed in the thickness direction (X direction) of the reference beam L, but the reference beam L may be formed as a spot beam having a predetermined divergence angle Ψ in any direction. When forming a spot beam, the beam shaping lens 22 is formed as a convex lens, and at least one of the entrance surface 22a and the exit surface 22b is formed as a convex surface that is convex along the X direction and the Y direction. This allows the beam shaping lens 22 to converge the reference beam L in the X direction and the Y direction to form a spot beam having a predetermined divergence angle Ψ. Alternatively, the beam shaping lens 22 is formed as a concave lens, and at least one of the entrance surface 22a and the exit surface 22b is formed as a concave surface that is concave along the X direction and the Y direction. This allows the beam shaping lens 22 to diffuse the reference beam L in the X direction and the Y direction to form a spot beam having a predetermined divergence angle Ψ. The reference light L of the present invention is not limited to a beam shape such as a slit light or a spot light, and is technically significant in that it is a diffused light having a predetermined diffusion angle Ψ.

[0063] 12A to 12C and the modified examples in FIGS. 13A to 13C, the thickness of the reference light L on the object W can be dynamically changed by moving the beam shaping lens 22 in the optical axis direction. Applying this to Equation 6 above, if the thickness of the reference light L on the object W (i.e., the width w of the image of the reference light L on the light-receiving surface) is multiplied by n, the angular velocity ω of the reference light L can be adjusted by n times, within the constraints of the high-speed performance of the image sensor 31. In other words, by adjusting the angular velocity ω of the reference light L by n times while multiplying the width w of the image of the reference light L on the light-receiving surface by n times, the required passing time T required for the image of the reference light L to pass a certain pixel on the light-receiving surface can be kept approximately constant, thereby enabling three-dimensional measurement with the required passing time T kept at or below the design lower limit Tmin.

[0064] A light projecting unit 2 including a moving mechanism 24 that moves the beam shaping lens 22 in the optical axis direction will be described. FIGS. 14A to 14C are top views of a light projecting unit 2 according to a further modification. The moving mechanism 24 includes an actuator 24a configured with an electric motor or the like, a rotary shaft 24b that can be rotated by the actuator 24a, and a moving unit 24c configured with a table or the like that can move in accordance with the movement of the rotary shaft 24b. The beam shaping lens 22 is fixed to the moving unit 24c. Viewed in the order of FIGS. 14A to 14C, it can be seen that the moving mechanism 24 moves the beam shaping lens 22 toward the light source 20, while viewed in the order of FIGS. 14C to 14A, it can be seen that the moving mechanism 24 moves the beam shaping lens 22 toward the object W. The thickness of the reference light L on the object W (i.e., the width w of the image of the reference light L on the light-receiving surface) can be adjusted by moving the beam shaping lens 22 in the optical axis direction.

[0065] By being able to adjust the thickness of the reference light L on the object W (i.e., the width w of the image of the reference light L on the light-receiving surface), it is possible to adjust the moving speed v of the image of the reference light L on the light-receiving surface and adjust the imaging time τ while keeping the passing time T required for the image of the reference light L to pass through a certain pixel on the light-receiving surface roughly constant, so that the cycle time of the entire system can be adjusted, for example, in accordance with the requirements of a system using the three-dimensional measuring device 1.

[0066] 12A to 12C, the light projecting unit 2 may be provided with a moving mechanism 24 that moves the collimating lens 21 in the optical axis direction, rather than moving the beam shaping lens 22 in the optical axis direction. In this case, the moving mechanism 24 includes an actuator 24a configured with an electric motor or the like, a rotating shaft 24b that can be rotated by the actuator 24a, and a moving unit 24c configured with a table or the like that can move in response to the operation of the rotating shaft 24b, and the collimating lens 21 is fixed to the moving unit 24c.

[0067] 12A to 12C, the light projecting unit 2 may include a moving mechanism 24 that moves the light source 20 in the optical axis direction instead of moving the collimator lens 21 in the optical axis direction. In this case, the moving mechanism 24 includes an actuator 24a configured with an electric motor or the like, a rotating shaft 24b that can be rotated by the actuator 24a, and a moving unit 24c configured with a table or the like that can move in response to the operation of the rotating shaft 24b, and the light source 20 is fixed to the moving unit 24c.

[0068] The configuration of a three-dimensional measuring device 1 that controls the angular velocity ω of the reference light L and the width w of the image of the reference light L will be described below. FIG. 15 is a block diagram of the three-dimensional measuring device 1 of one embodiment. The three-dimensional measuring device 1 includes a light-projecting unit 2 that projects reference light L having a predetermined diffusion angle Ψ onto an object W while scanning the object W at a predetermined angular velocity ω, a light-receiving unit 3 that receives the reference light L reflected from the object W, and a control unit 4 that controls the light-projecting unit 2 and the light-receiving unit 3. The three-dimensional measuring device 1 also includes a three-dimensional information calculation unit 5 that calculates three-dimensional information of the object W by triangulation based on information from the light-receiving unit 3.

[0069] Although not shown, the control unit 4 includes a drive circuit that controls the driving of the light projecting unit 2 and the light receiving unit 3. The control unit 4 may control the angular velocity ω of the scanning unit 23 in the light projecting unit 2 and the position in the optical axis direction of the beam shaping lens 22 in the light projecting unit 2 (i.e., the width w of the image of the reference light L on the light receiving surface). The angular velocity ω of the reference light L and the width w of the image of the reference light L are controlled based on the above-mentioned equation 6 so that the passing time T required for the light to pass through a certain pixel on the light receiving surface is approximately constant. This makes it possible to perform three-dimensional measurement under approximately constant conditions.

[0070] The three-dimensional information calculation unit 5 is configured by a computer equipped with a processor, memory, input / output interface, etc. The three-dimensional information calculation unit 5 sends the calculated three-dimensional information to an external device 6. The external device 6 is configured by another computer device, such as a control device that controls machines such as robots and machine tools, or a host computer device.

[0071] The external device 6 controls at least one of the position and posture of a control target portion of the machine using the three-dimensional information acquired from the three-dimensional measuring device 1. Alternatively, the external device 6 performs various simulations using the three-dimensional information acquired from the three-dimensional measuring device 1. In other embodiments, the three-dimensional measuring device 1 and the external device 6 may be integrated into one unit, or the three-dimensional measuring device 1 may be configured as various control devices, a host computer device, or the like.

[0072] According to the above embodiment, the reference light L is diffused light having a predetermined diffusion angle Ψ, so that a three-dimensional measuring device 1 capable of performing three-dimensional measurement under substantially constant conditions regardless of the distance to the object W can be provided.

[0073] The programs executed by the aforementioned processors, drive circuits, etc. may be provided by being recorded on a computer-readable non-temporary recording medium, such as a CD-ROM, or may be distributed via wired or wireless connections from a server device on a WAN (wide area network) or LAN (local area network).

[0074] Although various embodiments have been described herein, it should be recognized that the present invention is not limited to the above-described embodiments, but can be modified in various ways within the scope of the claims. [Explanation of symbols]

[0075] 1. Three-dimensional measurement device 2 Light emitter 3 Light receiving section 20 light source 21-22 Projection optical system 21 Collimating lens 21a Incidence plane 21b Output surface 22 Beam shaping lens 22a Incidence plane 22b Output surface 23 Scanning unit 24 Moving mechanism 24a actuator 24b Rotation axis 24c Moving part 30 Light receiving optical system 31 Image Sensor B Baseline length C center line D parallax f focal length I1~I4 images i Illuminance L reference light P, P1, P2 points on the object R Distance measurement range S Reference beam image T Time required for passage τ Imaging time t, t1~t3 time v Movement speed W Object w Width of the reference beam on the light receiving surface Z Distance to point P of the object θ Projection angle ω angular velocity φ Field of view Ψ Diffusion angle

Claims

1. a light projection unit that projects a reference light onto an object while scanning the object at a predetermined angular velocity; a light receiving unit that receives the reference light reflected from the object; a three-dimensional information calculation unit that calculates three-dimensional information of the object by triangulation based on information from the light receiving unit; Equipped with the reference light is diffused light having a predetermined diffusion angle, the light receiving unit includes a plurality of pixels, detects a change in illuminance when the reference light is illuminating the object for each pixel, and outputs the position of the pixel where the change in illuminance occurred, the time of the change, and the polarity of the change as event information.

2. 2. The three-dimensional measuring device according to claim 1, wherein the divergence angle and the angular velocity of the reference beam are determined so that a time required for the image of the reference beam to pass through a pixel of the light receiving unit is approximately constant.

3. The three-dimensional measuring apparatus according to claim 2 , wherein the required transit time is proportional to the divergence angle and inversely proportional to the angular velocity.

4. The three-dimensional measuring apparatus according to claim 1 , wherein the reference light is a slit light having the diffusion angle in a thickness direction.

5. 5. The three-dimensional measuring device according to claim 1, wherein the light-projecting unit comprises a beam-shaping lens that is convex along a first direction orthogonal to an optical axis of the reference light and that is concave along a second direction orthogonal to the optical axis and the first direction.

6. 6. The three-dimensional measuring device according to claim 1, wherein the light-projecting unit comprises a beam shaping lens that converges the reference light in a first direction perpendicular to an optical axis of the reference light to give the reference light the diffusion angle, and diffuses the reference light in a second direction perpendicular to the optical axis and the first direction to form a slit beam.

7. 5. The three-dimensional measuring device according to claim 1, wherein the light-projecting unit comprises a beam-shaping lens that is concave along a first direction orthogonal to an optical axis of the reference light and that is concave along a second direction orthogonal to the optical axis and the first direction.

8. 8. The three-dimensional measuring device according to claim 1, wherein the light-projecting unit comprises a beam shaping lens that diffuses the reference light in a first direction perpendicular to an optical axis of the reference light to give the reference light the diffusion angle, and diffuses the reference light in a second direction perpendicular to the optical axis and the first direction to form a slit beam.

9. The three-dimensional measuring device according to claim 1 , wherein the light projecting unit further comprises a movement mechanism that moves a lens or a light source in the optical axis direction.

Images