Surface roughness measuring device and surface roughness measuring method

Abstract

The present invention provides a surface roughness measuring device 100 which reduces measurement errors that arise due to reflection, by another illuminating portion, of light that has been specularly reflected at a measurement target surface, and which uses an illuminance difference stereo method to measure the surface roughness of the measurement target surface, wherein the surface roughness measuring device 100 is provided with an imaging unit 2 for imaging the measurement target surface W, an optical lens 3 for causing the imaging unit 2 to form an image of the measurement target surface W, and three of more illuminating portions 4a to 4e for emitting light onto the measurement target surface W, and wherein the illuminating portions 4a to 4e are disposed in positions not facing one another when viewed from the direction of the optical axis of the optical lens 3.

Description

Technical Field

[0001] The present invention relates to a surface roughness measuring device and a surface roughness measuring method.

Background Art

[0002] Conventionally, as a device for measuring the height, shape, etc. of a surface, as shown in Patent Document 1, a three-dimensional measuring device that measures the shape of the surface of a measurement object using the illuminance difference stereo method has been considered. This three-dimensional measuring device includes an imaging unit having an illumination unit and an imaging device. The illumination unit has a dome-shaped dome member with an opening formed at its top, and three or more LEDs (specifically, eight LEDs) arranged inside the dome member.

[0003] However, when the above three-dimensional measuring device is installed close to the measurement target surface, light from one LED is (macroscopically) specularly reflected on the measurement target surface, and the specularly reflected light reaches another LED arranged opposite. The specularly reflected light that reaches another LED is reflected by that another LED and enters the measurement target surface again. As a result, it causes a measurement error in the measurement of surface roughness using the illuminance difference stereo method.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] Therefore, the present invention has been made to solve the above-described problems, and its main problem is to reduce the measurement error caused by specularly reflected light on the measurement target surface being reflected by another illumination unit.

Means for Solving the Problems

[0006] In other words, the surface roughness measuring device according to the present invention is a surface roughness measuring device that measures the surface roughness of a surface to be measured using the illuminance difference stereo method, and comprises an imaging unit that images the surface to be measured, an optical lens that images the surface to be measured onto the imaging unit, and three or more illumination units that irradiate the surface to be measured with light, wherein each of the illumination units is positioned so as not to face each other when viewed from the optical axis direction of the optical lens.

[0007] In this type of surface roughness measuring device, each illumination unit is positioned so that it does not face each other when viewed from the optical axis direction of the optical lens. This prevents specularly reflected light from the surface being measured from reaching other illumination units. As a result, measurement errors caused by specularly reflected light from the surface being measured being reflected by other illumination units can be reduced.

[0008] As for the specific arrangement of the lighting units, it is conceivable that they are arranged so that the plane containing the optical axis of each lighting unit and the optical axis of the optical lens is not shared by each lighting unit.

[0009] More specifically, it is desirable that the illumination units be three or more odd numbers and arranged at equal intervals around the optical axis of the optical lens. With this configuration, each illumination unit will be positioned so that it does not face each other when viewed from the direction of the optical axis of the optical lens.

[0010] In a specific embodiment of the surface roughness measuring device of the present invention, it is conceivable to further include a lighting holding member that holds the plurality of lighting units and surrounds the surface to be measured in order to block the intrusion of light from the surrounding environment. In this configuration, in order to improve the accuracy of surface roughness measurement, it is desirable that the inner surface of the lighting holding member be treated with an anti-reflective coating.

[0011] The surface roughness measuring device using the illuminance difference stereo method further includes a calculation unit that calculates the normal vector of the surface to be measured at each pixel position from the brightness change at the same pixel position, i.e., the brightness information of each image, in a plurality of images captured by the imaging unit while switching between the plurality of illumination units. In this case, if specularly reflected light from the object to be measured is incident on the imaging unit, a portion of the image captured by the imaging unit will be overexposed, resulting in a measurement error. To solve this problem, it is desirable that the calculation unit calculates the normal vector by omitting the brightness information of images that have brightness values ​​exceeding a predetermined brightness.

[0012] As described above, in order to simplify the device configuration by omitting the brightness information of images with brightness values ​​exceeding a predetermined brightness and calculating the normal vector, it is desirable that there be five illumination units.

[0013] Furthermore, it is desirable that the surface roughness measuring device according to the present invention further comprises a control unit that controls the switching of the illumination of the plurality of illumination units and controls the imaging operation of the imaging unit, a calculation unit that calculates the surface roughness of the surface to be measured from a plurality of images captured by the imaging unit while switching the plurality of illumination units, and a display unit that displays the image captured by the imaging unit or the surface roughness calculated by the calculation unit.

[0014] Furthermore, the surface roughness measurement method according to the present invention is a surface roughness measurement method that measures the surface roughness of a surface to be measured using the illuminance difference stereo method, and is characterized in that it uses an imaging unit that images the surface to be measured, an optical lens that images the surface to be measured onto the imaging unit, and three or more illumination units that irradiate the surface to be measured with light, and each of the illumination units is positioned so that they do not face each other when viewed from the optical axis direction of the optical lens.

[0015] Furthermore, in the 3D measuring device described in Patent Document 1, if the 3D measuring device is installed close to the surface to be measured, the light from the illumination unit can no longer be considered parallel light. As a result, the direction of the light vector (illumination direction vector) arriving from the illumination unit changes depending on the position of the surface to be measured. Consequently, errors increase when calculating the normal vector in the illuminance difference stereo method.

[0016] Therefore, the main objective of the surface roughness measuring device of the present invention is to accurately measure surface roughness by taking into account the illumination direction vector at each position on the surface to be measured.

[0017] In other words, the surface roughness measuring device according to the present invention is a surface roughness measuring device that measures the surface roughness of a surface to be measured using the illuminance difference stereo method, comprising: an imaging unit that images the surface to be measured; an optical lens that causes the surface to be measured to form an image on the imaging unit; a plurality of illumination units that irradiate the surface to be measured with light; and a calculation unit that calculates the normal vector of the surface to be measured from a plurality of images captured by the imaging unit while switching between the plurality of illumination units, wherein the calculation unit calculates the normal vector by using a vector from each illumination unit toward the position of the surface to be measured corresponding to each pixel as the illumination direction vector at each pixel.

[0018] With this type of surface roughness measuring device, the normal vector is calculated by using the vector from each illumination unit to the position of the surface to be measured corresponding to each pixel as the illumination direction vector for each pixel. Therefore, surface roughness can be measured accurately regardless of changes in the illumination direction vector at the position of the surface to be measured.

[0019] Furthermore, the distance from each illumination unit to the position of the measurement target surface corresponding to each pixel, i.e., the length of the illumination direction vector, also changes according to the position of the measurement target surface corresponding to each pixel. It is desirable to correct the brightness information of each pixel according to this distance from each illumination unit and calculate the normal vector.

[0020] In addition, the illuminance distribution characteristics of each lighting unit also become error factors when calculating the normal vector. Therefore, it is desirable that the calculation unit corrects the luminance information of each pixel based on the illuminance distribution characteristics of each lighting unit and calculates the normal vector.

[0021] Furthermore, the peripheral light reduction characteristics of the optical lens also become error factors when calculating the normal vector. Therefore, it is desirable that the calculation unit corrects the luminance information of each pixel based on the peripheral light reduction characteristics of the optical lens and calculates the normal vector.

[0022] When, for example, a color CCD is used for the imaging unit, depending on the spectral characteristics of the lighting unit, only a specific color component may be likely to overexpose. Therefore, it is desirable that the calculation unit decomposes the color of the image captured by the imaging unit to generate a color-separated image and calculates the normal vector from the color-separated image based on the spectral characteristics of the lighting unit.

[0023] Specifically, it is desirable that the lighting unit uses white LEDs.

[0024] In white LEDs, the blue component may be extremely strong, and the blue component is likely to overexpose. Therefore, it is desirable that the calculation unit decomposes the color of the image captured by the imaging unit to generate a color-separated image and calculates the normal vector using the green image and / or the red image among the color-separated images.

[0025] Since the blue component is strong in white LEDs, it is desirable that the surface roughness measuring device of the present invention includes a control unit that adjusts the exposure of the imaging unit based on the blue component.

[0026] Moreover, it is desirable that the surface roughness measuring device according to the present invention further includes a control unit that controls the switching of the lighting of the plurality of lighting units and the imaging operation of the imaging unit, a calculation unit that calculates the surface roughness of the measurement target surface from a plurality of images captured by the imaging unit while switching the plurality of lighting units, and a display unit that displays the image captured by the imaging unit or the surface roughness calculated by the calculation unit.

[0027] Furthermore, the surface roughness measurement method according to the present invention is a surface roughness measurement method for measuring the surface roughness of a measurement target surface using the illuminance difference stereo method, including an imaging unit for imaging the measurement target surface, an optical lens for forming an image of the measurement target surface on the imaging unit, and a plurality of illumination units for irradiating the measurement target surface with light. While switching the plurality of illumination units, a normal vector of the measurement target surface is calculated from a plurality of images captured by the imaging unit. The normal vector is calculated by using, as an illumination direction vector at each pixel, a vector from each illumination unit toward the position of the measurement target surface corresponding to each pixel.

[0028] Moreover, the surface roughness measurement program according to the present invention is a program for measuring the surface roughness of a measurement target surface using the illuminance difference stereo method. It is used in a surface roughness measurement apparatus including an imaging unit for imaging the measurement target surface, an optical lens for forming an image of the measurement target surface on the imaging unit, and a plurality of illumination units for irradiating the measurement target surface with light. A computer is caused to have a function as a calculation unit that calculates a normal vector of the measurement target surface from a plurality of images captured by the imaging unit while switching the plurality of illumination units. The calculation unit calculates the normal vector by using, as an illumination direction vector at each pixel, a vector from each illumination unit toward the position of the measurement target surface corresponding to each pixel.

[0029] Furthermore, in the three-dimensional measurement apparatus of Patent Document 1 described above, a moving mechanism for moving the imaging unit and a transport unit for transporting the measurement object are required, resulting in a large-scale apparatus configuration and making it difficult to use at a processing site or a manufacturing site. Also, for a large measurement object that cannot be mounted on the transport unit, it was impossible to measure the surface roughness itself.

[0030] Therefore, the main problems of the surface roughness measurement apparatus of the present invention are to easily measure the surface roughness without damaging the measurement target surface and to achieve miniaturization and weight reduction.

[0031] In other words, the surface roughness measuring device according to the present invention is a surface roughness measuring device that measures the surface roughness of a surface to be measured using the illuminance difference stereo method, and comprises an imaging unit that images the surface to be measured, an optical lens that causes the surface to be measured to form an image on the imaging unit, a plurality of illumination units that irradiate the surface to be measured with light, and an illumination holding member that holds the plurality of illumination units, wherein the illumination holding member defines the distance between the optical lens and the surface to be measured.

[0032] With this type of surface roughness measuring device, the illumination holding member, which holds multiple illumination units, defines the distance between the optical lens and the surface to be measured, thus eliminating the need for a positioning mechanism to position the optical lens relative to the surface to be measured. As a result, the configuration of the surface roughness measuring device can be simplified, enabling miniaturization and weight reduction. Furthermore, since the surface roughness of the surface to be measured can be measured simply by applying the illumination holding member to the surface to be measured, the surface roughness measuring device is easy to handle, allowing for easy measurement of surface roughness in processing and manufacturing sites. Here, because the illuminance difference stereo method is used, the measurement time for the surface roughness of the surface to be measured can be shortened to, for example, about 1 second.

[0033] In order to enable the measurement of surface roughness of various target surfaces, it is desirable that the surface roughness measuring device of the present invention be configured so that the measurement range of surface roughness can be changed by replacing the illumination holding member.

[0034] A specific embodiment of the surface roughness measuring device of the present invention may further include a device body having the imaging unit and the optical lens. In this configuration, it is desirable that the illumination holding member be detachably attached to the main body of the device and define the distance between the optical lens and the surface to be measured. It is also desirable to appropriately set the imaging magnification of the optical lens according to the surface roughness measurement range. Furthermore, it is desirable that the distance between the optical lens and the surface to be measured, defined by the illumination holding member, be set according to the imaging magnification of the optical lens. In addition, it is desirable that the main body of the device and the illumination holding member be electrically connectable in order to supply power to multiple illumination units.

[0035] In order to accurately measure surface roughness using the illuminance difference stereo method, the positions of multiple illumination units relative to the imaging unit must be set in a predetermined position. In other words, the illumination holding member must be positioned relative to the main body of the device. To facilitate this positioning, it is desirable that the surface roughness measuring device of the present invention be equipped with a positioning mechanism that positions the illumination holding member around the optical axis of the optical lens relative to the main body of the device.

[0036] In order to accurately measure surface roughness using the illuminance difference stereo method, it is desirable that the optical axis of each of the multiple illumination units passes through the intersection point of the optical axis of the optical lens and the surface to be measured.

[0037] In order to automatically perform surface roughness measurement suitable for a given lighting holder when the lighting holder is replaced, it is desirable to further include a control unit that switches the surface roughness measurement range according to the type of lighting holder, changes the imaging magnification by the optical lens, and automatically changes the cutoff frequency when calculating surface roughness.

[0038] When the imaging magnification by the optical lens is high, the amount of light captured by the imaging unit decreases. Therefore, it is desirable that the illumination holding member where the imaging magnification by the optical lens is low uses a wide-angle LED as the illumination unit, and that the illumination holding member where the imaging magnification by the optical lens is high uses a narrow-angle LED as the illumination unit.

[0039] Furthermore, it is desirable that the surface roughness measuring device of the present invention further comprises a control unit that controls the switching of the illumination of the plurality of illumination units and controls the imaging operation of the imaging unit, a calculation unit that calculates the surface roughness of the surface to be measured from a plurality of images captured by the imaging unit while switching the plurality of illumination units, and a display unit that displays the images captured by the imaging unit or the surface roughness calculated by the calculation unit.

[0040] Furthermore, the surface roughness measurement method according to the present invention is a surface roughness measurement method that measures the surface roughness of a surface to be measured using the illuminance difference stereo method, and is characterized in that it uses an imaging unit that images the surface to be measured, an optical lens that images the surface to be measured onto the imaging unit, a plurality of illumination units that irradiate the surface to be measured with light, and an illumination holding member that holds the plurality of illumination units, and the distance between the optical lens and the surface to be measured is defined by the illumination holding member. [Effects of the Invention]

[0041] According to the present invention configured in this way, it is possible to reduce measurement errors caused by specular reflection of light from the surface to be measured being reflected by another illumination unit. [Brief explanation of the drawing]

[0042] [Figure 1] This is a schematic diagram showing the configuration of the surface roughness measuring device according to this embodiment. [Figure 2] This is a cross-sectional perspective view showing the optical axis, illumination section, and illumination holding member of the optical lens of the same embodiment. [Figure 3] This is a schematic plan view showing the arrangement of the lighting unit with respect to the lighting holding member in the same embodiment. [Figure 4] This diagram schematically shows the optical path of specularly reflected light when two illumination units are positioned opposite each other. [Figure 5] This is a schematic diagram showing the state of the lighting support member before and after replacement in the same embodiment. [Figure 6] This is a schematic cross-sectional view showing the configuration of the positioning mechanism of the same embodiment. [Figure 7]This figure shows the measurement results of the three-dimensional shape and surface roughness of the embodiment. [Figure 8] This is a schematic diagram illustrating the calculation method of the normal vector calculation unit in the same embodiment. [Figure 9] (a) The calculation results of the 3D shape when all brightness information is used, and (b) when pixels with brightness values ​​exceeding a predetermined brightness are excluded. [Figure 10] This is a schematic diagram showing the illumination direction vector from the illumination unit to the surface being measured. [Figure 11] (a) The calculation results of the 3D shape when the illumination direction vector is not corrected and (b) when the illumination direction vector is corrected for each pixel. [Figure 12] This figure shows the three-dimensional brightness distribution of each color component image when using a white LED. [Explanation of symbols]

[0043] 100... Surface roughness measuring device W... Surface to be measured 2. Imaging Unit 3. Optical lenses 3L... Optical axis of an optical lens 4a~4e...Multiple lighting units 4L... Optical axis of the lighting unit P... The intersection of the optical axis of the optical lens and the surface to be measured. 5. Lighting support component 6. Control Unit 7. Calculation Section 9. Main unit of the device 10. Positioning mechanism [Modes for carrying out the invention]

[0044] Hereinafter, one embodiment of the surface roughness measuring device according to the present invention will be described with reference to the drawings.

[0045] <Device configuration> The surface roughness measuring device 100 of this embodiment measures the surface roughness of the surface W to be measured using the illuminance difference stereo method, and is a portable device that can be moved relative to the surface W to be measured.

[0046] The illuminance difference stereo method is a technique that acquires three or more images of the surface to be measured W using a single imaging unit that is stationary relative to the surface to be measured W, while switching between multiple illumination units. From the brightness changes at the same pixel position in the three or more images, the normal vector of the surface to be measured at the position of the surface to be measured W corresponding to each pixel is obtained, and the height information of the surface to be measured is obtained by integrating the two-dimensional distribution of these normal vectors.

[0047] Specifically, as shown in Figure 1, the surface roughness measuring device 100 includes an imaging unit 2 that images the surface to be measured W, an optical lens 3 that images the surface to be measured W onto the imaging unit 2, a plurality of illumination units 4a to 4e that irradiate the surface to be measured W with light, an illumination holding member 5 that holds the plurality of illumination units 4a to 4e, a control unit 6 that controls the switching of the illumination of the plurality of illumination units 4a to 4e and controls the imaging operation of the imaging unit 2, and a calculation unit 7 that calculates the surface roughness of the surface to be measured W from a plurality of images captured by the imaging unit 2 while switching the plurality of illumination units 4a to 4e.

[0048] The imaging unit 2 captures images of the surface to be measured W and has an image sensor 21, such as a CCD image sensor or a CMOS image sensor. When capturing images of the surface to be measured W, the imaging unit 2 is positioned vertically above the surface to be measured W.

[0049] The optical lens 3 forms an optical image of the surface to be measured W onto the image sensor 21 of the imaging unit 2. This optical lens 3 is movable so that the distance between it and the image sensor 21 of the imaging unit 2 can be changed by a focusing mechanism (not shown). Furthermore, the optical axis 3L of the optical lens 3 is positioned by the illumination holding member 5 so that it is approximately perpendicular to the surface to be measured W when imaging the surface to be measured W.

[0050] The imaging unit 2 and optical lens 3 are housed in the housing 8 to constitute the main body of the device 9. This main body of the device 9 is equipped with a power switch and a measurement button to start measuring surface roughness. The control unit 6 and calculation unit 7 are connected to the main body of the device 9 via wired or wireless communication and consist of an information processing device COM, such as a computer, which is configured to display various information on a display unit 12, such as a display. The information processing device COM has a CPU, internal memory, input / output interface, etc., and performs the following functions based on the surface roughness measurement program stored in the internal memory.

[0051] Each of the multiple illumination units 4a to 4e is an LED, and in this embodiment, white LEDs are used. Furthermore, as shown in Figure 2, the optical axis 4L of each of the multiple illumination units 4a to 4e is positioned to pass through the intersection point P between the optical axis 3L of the optical lens 3 and the surface to be measured W when imaging the surface to be measured W. Since the multiple illumination units 4a to 4e have an axially symmetric illuminance distribution with respect to their respective optical axes 4L, the illuminance within the measurement range of the surface to be measured W can be made uniform.

[0052] The illumination holding member 5, which holds these multiple illumination units 4a to 4e, has a roughly rotating shape, and the light-emitting surfaces of the illumination units 4a to 4e are exposed on its inner surface. In this embodiment, as shown in Figure 3, five illumination units 4a to 4e are provided on the illumination holding member, and these illumination units 4a to 4e are provided at equal intervals around the optical axis 3L of the optical lens 3.

[0053] In this configuration, if any of the illumination units 4a to 4e are facing each other, as shown in Figure 4, light from one illumination unit is specularly reflected (macroscopically) from the surface W to be measured, and this specularly reflected light reaches the other illumination unit positioned opposite it. The specularly reflected light that reaches the other illumination unit is reflected again from the light-emitting surface of that illumination unit and enters the surface W to be measured again. As a result, this causes measurement errors in surface roughness measurement using the illuminance difference stereo method.

[0054] Therefore, in order to reduce measurement errors caused by specular reflection of the surface W being measured being reflected by another illumination unit, the multiple illumination units 4a to 4e are arranged in positions that do not face each other when viewed from the optical axis direction of the optical lens 3, as shown in Figure 3.

[0055] In this configuration, each of the illumination units 4a to 4e is positioned so that they do not face each other when viewed from the optical axis direction of the optical lens 3. This prevents specularly reflected light from the surface W being measured from reaching other illumination units in each of the illumination units 4a to 4e. As a result, measurement errors caused by specularly reflected light from the surface W being reflected by other illumination units can be reduced.

[0056] Specifically, the planes including the optical axis 4L of each illumination unit 4a to 4e and the optical axis 3L of the optical lens 3 are arranged so that they are not shared by each illumination unit 4a to 4e (see Figure 2). In this embodiment, the five illumination units 4a to 4e are arranged at equal intervals around the optical axis 3L of the optical lens 3 (see Figure 3).

[0057] As shown in Figure 1, the illumination holding member 5 has its upper end fixed to the main body of the device 9, and its lower end contacts the surface to be measured W, thereby defining the distance between the optical lens 3 and the surface to be measured W. The illumination holding member 5 has a roughly revolutionary shape so as to surround the area centered on the intersection point P between the optical axis 3L of the optical lens 3 and the surface to be measured W, so that ambient light (such as light from indoor lighting equipment like fluorescent lamps or sunlight) does not enter the measurement range of the surface to be measured W. The inner surface of this illumination holding member 5 may be treated to prevent reflection, for example, by painting it with a material with low reflectivity or by attaching a sheet with low reflectivity.

[0058] With this configuration, the illumination holding member 5, which holds multiple illumination units 4a to 4e, defines the distance between the optical lens 3 and the surface to be measured W, thus eliminating the need for a positioning mechanism to position the optical lens 3 relative to the surface to be measured W. As a result, the configuration of the surface roughness measuring device 100 can be simplified, enabling miniaturization and weight reduction. Furthermore, since the surface roughness of the surface to be measured W can be measured simply by placing the illumination holding member 5 over the surface to be measured W, the surface roughness measuring device 100 is easy to handle, allowing for easy measurement of surface roughness in processing and manufacturing sites. Here, because the illuminance difference stereo method is used, the measurement time for the surface roughness of the surface to be measured W can be shortened to, for example, less than 1 second.

[0059] Furthermore, as shown in Figure 5, the surface roughness measuring device 100 of this embodiment is configured to allow the measurement range of surface roughness to be changed by replacing the illumination holding member 5. Specifically, the illumination holding member 5 is detachably attached to the device body 9, and the measurement range of surface roughness can be changed by replacing the illumination holding member 5.

[0060] Here, we will explain the measurement range for surface roughness. For example, when measuring a surface W with relatively severe irregularities, such as tens of micrometers, the period of the irregularities is expected to be long. Therefore, a wider area must be measured. On the other hand, when measuring a smooth surface W with irregularities of the submicron scale, fine irregularities are expected to occur at extremely short intervals. Therefore, it is more efficient to narrow the measurement range and increase the positional resolution. In other words, it is necessary to widen or narrow the measurement range according to the order of magnitude of the surface roughness to be measured, and a reference length Lr is defined as the basis for this (for example, shown in Tables 1 and 2 of JIS B0633:2001). The reference length Lr changes in steps in response to increases or decreases in the surface roughness to be measured, so this is called the measurement range for surface roughness.

[0061] As described above, the optical lens 3 forms an optical image of the measurement range on the surface W to be measured onto the image sensor 21. Here, since the image sensor 21 has many pixels in two dimensions, vertically and horizontally, the measurement range that contributes to the captured image as an optical image is a rectangular region on the surface W to be measured. This rectangular region of the measurement range must include the aforementioned reference length Lr (at least the diagonal length of the rectangle must be longer than the reference length Lr). Since the size of the image sensor 21 is finite, the imaging magnification of the optical lens 3 must be set so that the size of this measurement range can be changed according to the measurement range of the surface roughness. For example, if the surface roughness to be measured is large, the aforementioned reference length Lr will be long. The imaging magnification of the optical lens 3 is set to a low magnification so that the measurement range is a rectangular region that includes this longer reference length Lr. Conversely, if the surface roughness to be measured is small, the aforementioned reference length Lr will be short, so the imaging magnification of the optical lens 3 is increased as much as possible to increase the positional resolution (shortening the distance on the surface W to be measured that corresponds to the distance between adjacent pixels).

[0062] Incidentally, the illumination holding member 5 defines the distance between the device body 9 (specifically the optical lens 3) and the surface to be measured W according to the measurement range. In this embodiment, when the imaging magnification of the optical lens 3 is high, the illumination holding member 5 that shortens the distance between the optical lens 3 and the surface to be measured W is used (see Figure 5(a)), and when the imaging magnification of the optical lens 3 is low, the illumination holding member 5 that lengthens the distance between the optical lens 3 and the surface to be measured W is used (see Figure 5(b)). This makes it possible to form an optical image of the surface to be measured W in focus on the image sensor 21, regardless of the imaging magnification of the optical lens 3.

[0063] In the illumination holding member 5, where the imaging magnification by the optical lens 3 is low, wide-angle LEDs (for example, with a half-angle of 60 degrees and a luminous intensity of 1.5 cd) are provided as illumination units 4a to 4e. This allows for uniform illumination over a wide measurement range. On the other hand, when the imaging magnification by the optical lens 3 is high, the F-number (the reciprocal of the brightness of the optical system) of the entire optical system including the optical lens 3 increases, resulting in a decrease in the amount of light reaching the image sensor 21. Therefore, narrow-angle LEDs (for example, with a half-angle of 30 degrees and a luminous intensity of 25 cd) are used as illumination units 4a to 4e in the illumination holding member 5. Narrow-angle LEDs have a higher luminous intensity, which is a measure of brightness as a light source, so they can compensate for the decrease in the amount of light reaching the image sensor 21. However, with narrow-angle LEDs, the half-angle (the angle in the radiation direction where the amount of light is halved, relative to the optical axis of the LED where the amount of light is maximum) is small, making them unsuitable for uniformly illuminating a wide area. However, when the magnification of the optical lens 3 is high, the measurement range becomes narrower, so uniform illumination is possible within the measurement range.

[0064] In this embodiment, as shown in Figure 6, a positioning mechanism 10 is provided to position the illumination holding member 5 around the optical axis 3L of the optical lens 3 relative to the main body 9 of the device.

[0065] This positioning mechanism 10 positions the illumination holding member 5 relative to the main body 9, thereby positioning multiple illumination units 4a to 4e around the optical axis 3L of the optical lens 3 with respect to the image sensor of the imaging unit 2.

[0066] Specifically, the positioning mechanism 10 has a positioning projection 101 provided on either the device body 9 or the lighting holding member 5, and a positioning recess 102 provided on the other of the device body 9 or the lighting holding member 5, into which the positioning projection 101 fits. In this embodiment, the positioning projection 101 is provided on the lighting holding member 5, and the positioning recess 102 is provided on the device body 9.

[0067] This allows the positions of each illumination unit 4a to 4e, as seen from the image sensor 21, to be returned to their pre-set correct positions, even when the illumination holding member 5 is attached or detached. The positions of each illumination unit 4a to 4e are necessary for the calculation of the normal vector described later, so if they change when the illumination holding member 5 is attached or detached, it will cause errors in the calculation of the normal vector. However, in this embodiment, a positioning mechanism 10 is provided, which makes it possible to reduce errors associated with attachment and detachment.

[0068] Furthermore, in this embodiment, as shown in Figure 6, the main body 9 is provided with a main body-side connector 111 for supplying power to multiple lighting units 4a to 4e, and the lighting holding member 5 is provided with a lighting-side connector 112 for supplying power to multiple lighting units 4a to 4e. By attaching the lighting holding member 5 to the main body 9, the main body-side connector 111 and the lighting-side connector 112 are connected. Specifically, the main body 9 and the lighting holding member 5 are positioned and attached by the positioning mechanism 10 described above, thereby connecting the main body-side connector 111 and the lighting-side connector 112. Since each lighting unit 4a to 4e is electrically connected independently, any lighting unit can be turned on or off by the control unit 6.

[0069] In this embodiment, in addition to the positioning mechanism 10 described above, a main unit-side connector 111 and a lighting-side connector 112 for electrical connection are provided separately. However, it is also acceptable to use only the main unit-side connector 111 and the lighting-side connector 112. In this case, the main unit-side connector 111 and the lighting-side connector 112 can be connected to perform the function of the positioning mechanism 10 described above.

[0070] Furthermore, the surface roughness measuring device 100 of this embodiment is configured to automatically switch the surface roughness measurement range according to the type of lighting holding member 5 by replacing the lighting holding member 5.

[0071] Specifically, when the main unit connector 111 and the illumination-side connector 112 are connected, the control unit 6 of the surface roughness measuring device 100 recognizes the type of illumination holding member 5. Then, it switches the surface roughness measurement range according to the type of illumination holding member 5 and selects the aforementioned reference length Lr. Furthermore, it changes the imaging magnification by the optical lens 3 and automatically changes the cutoff frequencies (λc, λs, λf) used when calculating surface roughness corresponding to the reference length Lr.

[0072] Next, the operation of the surface roughness measuring device 100 of this embodiment during surface roughness measurement will be described. When the control unit 6 detects that the user has pressed the measurement button provided on the device body 9, it turns on the illumination unit 4a and turns off the other illumination units 4b to 4e. In this state, the optical image formed by the optical lens 3 is recorded as image data by the image sensor 21 of the shooting unit 2. This image data is transferred to the calculation unit 7 of the information processing device COM. Next, the control unit 6 turns off the illumination unit 4a and turns on the illumination unit 4b instead, and transfers the image data captured by the image sensor 21 to the calculation unit 7. In this way, shooting is performed while switching to a state where only one of the multiple illumination units 4a to 4e is lit, and a total of 5 images are obtained, the same number as the illumination units 4a to 4e.

[0073] As shown in Figure 1, the calculation unit 7 includes a normal vector calculation unit 7a that calculates the normal vector of the surface to be measured W from multiple images captured by the imaging unit 2 while switching between illumination units 4a to 4e, a three-dimensional shape calculation unit 7b that calculates the three-dimensional shape of the surface to be measured W by integrating the normal vector of the surface to be measured W, and a surface roughness calculation unit 7c that calculates the surface roughness from the three-dimensional shape of the surface to be measured W. If the cutoff frequency is changed by the control unit 6, the surface roughness calculation unit 7c calculates the surface roughness (for example, the arithmetic mean roughness Ra of the roughness curve in JIS B0601:2013, the maximum height Rz of the roughness curve, etc.) using the changed cutoff frequency.

[0074] Specifically, the normal vector calculation unit 7a calculates the normal vector at each pixel position from the brightness change at the same pixel position in multiple images, i.e., from the brightness information of each image. In this embodiment, the normal vector calculation unit 7a calculates the normal vector from the brightness information of the same pixel position in five images and the illumination direction vector indicating the direction of the incident illumination when each image was captured, for example, using the least squares method.

[0075] The 3D shape calculation unit 7b determines the surface shape of the surface to be measured W by integrating the normal vector calculated by the normal vector calculation unit 7a. Figure 7 shows an example of the surface shape of the surface to be measured W (ceramic) determined by the 3D shape calculation unit 7b. The top of Figure 7 shows the height dimensions (roughness curves) of the surface to be measured W in two mutually orthogonal directions (X direction and Y direction). The 3D shape obtained by the 3D shape calculation unit 7b can be displayed on the display unit 12.

[0076] The surface roughness calculation unit 7c calculates the surface roughness of the surface to be measured W from the three-dimensional shape obtained by the three-dimensional shape calculation unit 7b. Here, the surface roughness can be, for example, the arithmetic mean roughness Ra or the maximum height Rz of the roughness curve according to JIS B0601:2013. The surface roughness obtained by the surface roughness calculation unit 7c can be displayed on the display unit 12 and can be displayed on the same screen as the three-dimensional shape (see the top of Figure 7).

[0077] In the surface roughness measuring device 100 described above, when the inclination direction and angle at a point on the surface to be measured W satisfy certain conditions, specular reflected light from the surface to be measured W is incident on the imaging unit 2 when a specific illumination unit among the multiple illumination units 4a to 4e is turned on. Since specular reflected light is orders of magnitude brighter than diffuse reflected light, the photoelectrically converted charge in the pixels of the image sensor 21 that receive this light becomes saturated. As a result, a part of the image captured by the imaging unit 2 may be overexposed and observed as a bright spot. In such cases, there is a risk of measurement errors occurring when calculating the normal vector. To solve this problem, the normal vector calculation unit 7a can calculate the normal vector by omitting the brightness information of images with brightness values ​​exceeding a predetermined brightness. Consider the case where the multiple images captured by the imaging unit 2 while switching between illumination units 4a to 4e are the five images shown in Figure 8. In the image taken with illumination unit 4b turned on, a bright spot can be seen as shown in the lower left. Therefore, when calculating the normal vector of the pixel corresponding to this bright spot, the brightness information 2 obtained from the image taken with the illumination unit 4b lit can be omitted, and the calculation can be performed using the brightness information 1, 3, 4, and 5 from the remaining four images. That is, the brightness information 1 to 5 for pixels at the same position in the five captured images is checked, and the brightness information of images with brightness values ​​exceeding a predetermined brightness is omitted, and the normal vector for that pixel position is calculated from the brightness information of the remaining images and the illumination direction vector. On the other hand, as shown in the upper right of Figure 8, if the brightness information 1 to 5 for pixels at the same position is below the predetermined brightness, the normal vector can be calculated using all the brightness information and the illumination direction vector. Figure 9(a) shows the calculation result of the 3D shape when all brightness information is used for all pixel positions. In contrast, Figure 9(b) shows the calculation result of the 3D shape when the brightness information of the same pixel is checked and the brightness information of images with brightness values ​​exceeding a predetermined brightness is omitted, and it can be seen that spike-like protrusions and holes are reduced.

[0078] Furthermore, since the optical axes 4L of the multiple illumination units 4a to 4e are tilted with respect to the measurement target surface W, if the measurement target surface W has a large protrusion, for example, shadows may be cast in the area adjacent to the protrusion. Similarly, concave areas such as large holes will also cast shadows. The shadowed areas are not visible from the image sensor 21, which leads to errors when calculating the normal vector. To solve this problem, the normal vector calculation unit 7a can calculate the normal vector by omitting the brightness information of images with brightness values ​​below a second predetermined brightness. That is, the brightness information 1 to 5 at the same position of pixels in the five captured images is checked, and the brightness information of images with brightness values ​​below the second predetermined brightness is omitted, and the normal vector at that pixel position is calculated from the brightness information of the remaining images and the illumination direction vector.

[0079] Furthermore, even without defining a predetermined brightness as described above, "measurement errors due to the influence of bright spots and shadows" can be avoided in the following way. When calculating the normal vector for each pixel, five brightness information values ​​can be used. Of these, if three or more brightness information values ​​are available, the normal vector can be calculated, so we consider all possible combinations. That is, there is one case where all five brightness information values ​​are used. There are five cases where four brightness information values ​​are used. And there are 10 combinations where any three brightness information values ​​are used. In total, the normal vector can be calculated in 16 combinations, so we can perform a predetermined error evaluation and select the combination that minimizes the measurement error, and then adopt the normal vector calculated using that combination. For error evaluation, for example, the brightness value can be inversely calculated using the calculated normal vector and the illumination direction vector used during the calculation, and the sum of the squares of the residuals with the brightness information of each image can be used.

[0080] Furthermore, in the surface roughness measuring device 100 of this embodiment, since the illumination units 4a to 4e are installed in close proximity to the surface to be measured, the light emitted from the illumination units 4a to 4e can no longer be considered parallel light. As a result, depending on the position of the surface W to be measured, the direction of the light vector (illumination direction vector) reaching each position in the measurement range from the illumination units 4a to 4e changes, as shown in Figure 10. Consequently, errors increase when calculating the normal vector in the illuminance difference stereo method.

[0081] Therefore, in order to accurately measure the surface roughness by considering the illumination direction vector at each position on the surface W to be measured, the normal vector calculation unit 7a can calculate the normal vector by using the vectors from each illumination unit 4a to 4e pointing to the position on the surface W to be measured corresponding to each pixel as the illumination direction vector for each pixel. Specifically, the following calculation is performed for each pixel. First, a three-dimensional coordinate system is considered with the intersection point P of the optical axis 3L of the optical lens 3 and the surface W to be measured as the origin. In this three-dimensional coordinate system, the position vector indicating the position of each illumination unit 4a to 4e is taken as V1. Then, the position on the surface W to be measured corresponding to the pixel is calculated for each pixel as the position vector V2 in the same three-dimensional coordinate system, and the difference between the two, vector V2-V1, is calculated and taken as the illumination direction vector.

[0082] With this configuration, the normal vector is calculated by using the vectors from each illumination unit 4a to 4e that point to the position of the measurement target surface W corresponding to each pixel as the illumination direction vector at each pixel. Therefore, the surface roughness can be measured accurately by taking into account the illumination direction vector at each position on the measurement target surface W. Figure 11 shows the results of measuring the three-dimensional shape of a flat plate, showing (a) when the illumination direction vector is not corrected (considered as parallel light) and (b) when the illumination direction vector is corrected for each pixel. It can be seen that the degree of curvature of the surface shape is reduced when the illumination direction vector is corrected for each pixel.

[0083] Furthermore, the distance from each illumination unit 4a to 4e to the position on the measurement target surface W corresponding to each pixel, i.e., the length of the illumination direction vector V2-V1, also changes according to the position on the measurement target surface corresponding to each pixel. In lighting equipment that emits light radially, not limited to LEDs, the illuminance, which is the brightness per unit area, decreases inversely proportional to the square of the reach. Therefore, it is desirable to correct the brightness information of each pixel and calculate the normal vector according to the distance from each illumination unit, i.e., the magnitude (scalar value) of the illumination vector V2-V1.

[0084] Furthermore, each illumination unit 4a to 4e has an axially symmetric illuminance distribution characteristic centered on the optical axis 4L. The aforementioned half-angle is a typical example, and these illuminance distribution characteristics can be stored in the calculation unit 7. It is desirable to use these illuminance distribution characteristics to correct the luminance information at each pixel position and calculate the normal vector.

[0085] Furthermore, the optical lens 3 has a peripheral vignetting characteristic. This refers to the phenomenon where the image becomes darker towards the corners compared to the center of the captured image. In other words, even if the actual illuminance distribution is uniform, the amount of light that passes through the optical lens 3 and reaches the image sensor 21 varies from pixel to pixel. Since this vignetting characteristic can also be stored in the calculation unit 7, it is desirable to correct the luminance information at each pixel position and calculate the normal vector.

[0086] Incidentally, when a user measures surface roughness, it is convenient to be able to sequentially observe (monitor) which position is being measured using images or other means. For this reason, it is desirable that when the power switch of the main unit 9 of the device is turned on, the images captured by the imaging unit 2 are continuously displayed on the display unit 12. The user can then move the main unit 9 to the appropriate position while looking at the images displayed on the display unit 12.

[0087] In this case, the image viewed by the user should preferably be a color image, and it is necessary to use an image sensor such as a color CCD or color CMOS as the image sensor 21. Furthermore, it is preferable that each illumination unit 4a to 4e be a white LED. By doing so, the user can move the main body of the device 9 to a position suitable for measuring surface roughness while observing an image consisting of correct color information.

[0088] When an image sensor such as a color CCD or color CMOS is used as the image sensor 21, one possible method is to convert the color image to monochrome, replace it with only brightness information, and then calculate the normal vector using the brightness information of the monochrome image. However, this is not always desirable. It is preferable for the normal vector calculation unit 7a to color-separate the image captured by the imaging unit 2 to generate a color-separated image, and then calculate the normal vector from the color-separated image based on the spectral characteristics of the illumination units 4a to 4e. The reasons for this will be explained below.

[0089] In the white LEDs used in each illumination section 4a to 4e, the blue component is generally extremely strong. Figure 12 shows the luminance distribution of each color-separated image in three dimensions, after the image, captured by a color CMOS image sensor while illuminating a smooth white plane with white LEDs, has been decomposed into blue (B), green (G), and red (R) color components. In the figures showing the B, G, and R color components, the magnitude of the luminance value is plotted vertically against the position on the plane XY corresponding to each pixel position. Looking at these figures, it can be seen that only the blue (B) component is out of range. In other words, the blue component is blown out.

[0090] Therefore, it is desirable for the normal vector calculation unit 7a to color-separate the image captured by the imaging unit 2 to generate a color-separated image, and then calculate the normal vector using the green image and / or red image from the color-separated image.

[0091] Furthermore, since white LEDs have a strong blue component, it is conceivable that the control unit 6 adjusts the exposure of the imaging unit 2 based on the blue component. As shown in Figure 12, the reason that only the blue component is out of range is that white LEDs have an extremely strong blue component, and exposure adjustment is performed based on the green component, which has the highest resolution for the human eye. When the user observes the measurement target surface W via the display unit 12, it is preferable to adjust the exposure based on the green component. However, when the control unit 6 detects that the measurement button on the main body 9 has been pressed and starts taking images to calculate the normal vector, it is desirable to perform exposure adjustment again based on the blue component.

[0092] <Effects of this embodiment> With such a surface roughness measuring device 100, the illumination units 4a to 4e are positioned so that they do not face each other when viewed from the optical axis direction of the optical lens 3. This prevents specularly reflected light from the surface to be measured W from reaching other illumination units in each illumination unit 4a to 4e. As a result, measurement errors caused by specularly reflected light from the surface to be measured W being reflected by other illumination units can be reduced.

[0093] <Other Embodiments> However, the present invention is not limited to the embodiments described above.

[0094] For example, although the above embodiment had a configuration with five lighting units, any configuration with three or more lighting units is acceptable.

[0095] Furthermore, although the above embodiment had an odd number of illumination units, a configuration with an even number of illumination units is also possible. Even in this case, it is desirable that each of the even number of illumination units be positioned so that they do not face each other when viewed from the optical axis direction of the optical lens 3.

[0096] Furthermore, in the above embodiment, the normal vector is calculated by using a vector from each illumination unit to the position of the measurement target surface corresponding to each pixel as the illumination direction vector for each pixel. However, the normal vector may also be calculated using a common illumination direction vector for each illumination unit.

[0097] Furthermore, although a white LED is used in the above embodiment, LEDs of other colors may also be used.

[0098] Furthermore, the normal vector calculation unit 7a may calculate the normal vector without performing color separation on the image captured by the imaging unit 2. Alternatively, if color separation is performed, the normal vector may be calculated using a color image other than the green image and / or red image. In addition, the exposure adjustment of the imaging unit 2 may be performed based on a color component other than the blue component.

[0099] Furthermore, it goes without saying that the present invention is not limited to the embodiments described above, and various modifications are possible without departing from its spirit. [Industrial applicability]

[0100] According to the present invention, measurement errors caused by specular reflection of light from the surface to be measured being reflected by another illumination unit can be reduced.

Claims

1. A surface roughness measuring device that measures the surface roughness of a surface to be measured using the illuminance difference stereo method, An imaging unit for imaging the surface to be measured, The imaging unit includes an optical lens for forming an image of the surface to be measured, Three or more illumination units that irradiate the surface to be measured with light, A rotating lighting holding member that holds the plurality of lighting units, The system comprises a housing that accommodates the optical lens and is fixed to the upper end of the illumination holding member, Each of the aforementioned illumination units is positioned on the inner surface of the illumination holding member so as not to face each other when viewed from the optical axis direction of the optical lens. A surface roughness measuring device wherein the illumination holding member has its lower end in contact with the surface to be measured, thereby defining the distance between the optical lens and the surface to be measured.

2. The surface roughness measuring device according to claim 1, wherein the plane including the optical axis of each illumination unit and the optical axis of the optical lens is arranged so as not to be shared by each illumination unit.

3. The surface roughness measuring device according to claim 1 or 2, wherein the illumination units are three or more odd numbers and are arranged at equal intervals around the optical axis of the optical lens.

4. The facility further comprises a lighting holding member that holds the plurality of lighting units and surrounds the surface to be measured, The surface roughness measuring device according to any one of claims 1 to 3, wherein the inner surface of the lighting holding member is treated with an anti-reflective coating.

5. The system further includes a calculation unit that calculates the normal vector of the measurement target surface at each pixel position from the brightness change at the same pixel position in multiple images captured by the imaging unit while switching between the multiple illumination units, The surface roughness measuring device according to any one of claims 1 to 4, wherein the calculation unit calculates the normal vector by omitting the brightness information of an image having a brightness value exceeding a predetermined brightness.

6. The surface roughness measuring device according to any one of claims 1 to 5, wherein the illumination unit comprises five units.

7. A control unit controls the switching of the illumination of the plurality of illumination units and controls the imaging operation of the imaging unit, A calculation unit that calculates the surface roughness of the surface to be measured from multiple images captured by the imaging unit while switching between the multiple illumination units, The surface roughness measuring device according to any one of claims 1 to 6, further comprising a display unit that displays an image captured by the imaging unit or a surface roughness calculated by the calculation unit.

8. The system includes a calculation unit that calculates the normal vector of the surface to be measured from multiple images captured by the imaging unit while switching between multiple illumination units, The surface roughness measuring apparatus according to any one of claims 1 to 7, wherein the calculation unit calculates the normal vector by using a vector from each illumination unit toward the position of the surface to be measured corresponding to each pixel as the illumination direction vector at each pixel.

9. The aforementioned lighting unit uses white LEDs. The surface roughness measuring device according to claim 8, wherein the calculation unit generates a color-separated image by color-separating the image captured by the imaging unit, and calculates the normal vector using the green image and / or red image from the color-separated image.

10. In the illumination holding member, where the imaging magnification by the optical lens is low, a wide-angle LED is used as the illumination unit. The surface roughness measuring device according to claim 1, wherein the illumination holding member, which enables high magnification imaging by the optical lens, uses a narrow-angle LED as the illumination unit.

11. A surface roughness measurement method for measuring the surface roughness of a surface to be measured using the illuminance difference stereo method, A surface roughness measurement method comprising: an imaging unit for imaging the surface to be measured; an optical lens for imaging the surface to be measured; three or more illumination units for irradiating the surface to be measured with light; a rotating illumination holding member for holding the illumination units; and a housing for housing the optical lens and fixed to the upper end of the illumination holding member, wherein each of the illumination units is positioned on the inner surface of the illumination holding member so as not to face each other when viewed from the optical axis direction of the optical lens, and the lower end of the illumination holding member is brought into contact with the surface to be measured to define the distance between the optical lens and the surface to be measured.

Images