JP2018132516A - Extraction device for diamond fine abrasive grain image - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an extraction device for a diamond fine abrasive grain image, which is integrated with a technique to selectively extract only a diamond fine abrasive grain image contributing to processing in performance analysis of a precision abrasive grain tool.SOLUTION: An extraction device has a function that obtains images taken by radiating light from an ultraviolet light wavelength region light source and a visible light wavelength region light source in order to extract a diamond fine abrasive grain image in an image obtained by photographing a tool work face of a precision abrasive grain tool, analyzes reflection light intensity ratios from objects on the images obtained under the respective light sources, and thereby extracts the diamond fine abrasive grain image.SELECTED DRAWING: Figure 1

Description

  The present invention provides, for example, a chip formed by sintering a material obtained by kneading a diamond fine abrasive grain that becomes a cutting edge at the time of grinding, a binder, and an abrasive layer additive such as carbon called a filler into a required shape. In a precision abrasive tool that performs grinding by placing and fixing on a base metal, by imaging the processing contribution surface of the chip and analyzing the image, the distribution of diamond fine abrasive grains and each diamond fine abrasive grain When trying to analyze the characteristics of the abrasive cutting edge, etc., the diamond fine grinding is performed from the images of the processing marks of the base metal, binder, filler, diamond fine abrasive grains, etc. The present invention relates to an apparatus for extracting a diamond fine abrasive grain image having a technique for selectively extracting a grain image.

  In recent years, it is called a resinoid bond diamond wheel in which a chip formed by sintering a material made by kneading fine diamond abrasive grains with a particle size of 50 μm to 300 μm, a binder and a filler into a required shape is placed and fixed on a base metal. Precision abrasive tools have been developed (see Non-Patent Document 1 and Non-Patent Document 2). In machining using such precision abrasive tools, the machining quality represented by the finished surface roughness accuracy of the workpiece work surface functions as an abrasive cutting edge protruding from the tool work surface of the precision abrasive tool. It is known that the abrasive cutting edge characteristics represented by the surface shape and size of diamond fine abrasive grains, the amount of protrusion of the abrasive grains, the arrangement of the abrasive grains, etc. affect the tool performance. Therefore, in order to perform high quality machining, it is important to accurately grasp the abrasive cutting edge characteristics of the tool work surface.

  With regard to accurately grasping the abrasive cutting edge characteristics of precision abrasive tools, the entire area of the tool work surface is imaged with a high-magnification imaging device, and all the abrasive particles that contribute to machining by analyzing this image A method for measuring and quantifying the abrasive cutting edge characteristics is established as a grinding surface inspection system for grinding tools or a grinding wheel cutting edge characteristics observation apparatus (see Patent Document 1 to Patent Document 4).

JP 2004-042158 A JP 2004-045078 A JP 2007-260857 A JP2013-002810A JP-A-11-352068 JP-A-64-0285544

"Evaluation of Grinding Performance by Mechanical Properties of Super Abrasive Wheels-1st Report: Estimation of Critical Abrasive Holding Force of Resinoid Bond Diamond Wheel-" Journal of the Agricultural Society of Japan, Vol. 159, no. 3, 2015, p134-p139 Evaluation of grinding performance by mechanical properties of superabrasive wheel-2nd report: Relationship between mechanical performance and grinding performance of abrasive layer of resinoid bond diamond wheel- 159, no. 4, 2015, p191-p196

  The analysis method used by the grinding surface inspection system proposed in Patent Document 1 to Patent Document 4 is a method in which fine abrasive grains such as diamond are regularly arranged and fixed to a base metal by electrodeposition or the like (an example of a processing contribution surface image) This is an effective technique for an aligned abrasive tool (shown in FIG. 4). However, in the case of an abrasive tool (for example, the above-mentioned resinoid bond diamond wheel) of a type in which diamond fine abrasive grains are kneaded with a binder and further solidified by sintering or the like with a filler mixed therein, the binder or filler There is also a thing that appears as a shape similar to diamond fine abrasive grains, and because the processing trace of the base metal is reflected as a shape confusing to the diamond fine abrasive grains, it can not be applied as a thing that requires sufficient analysis accuracy There are challenges. An example of a captured image of a processing-contributing surface in which diamond fine abrasive grains, a base metal, a binder, and a filler are mixed is shown in FIG. In this example, it is not clear whether the image shown in the form of particles is a diamond fine abrasive grain or something else.

  In Patent Document 5, for the purpose of discriminating between natural diamond and artificial diamond, an absorption characteristic near a wavelength of 415 nm in the visible light region, which is an event peculiar to natural diamond, is used, and a visible light wavelength region of I2 wavelength type or more is used. Visible light reflected on the top surface of the processed diamond in order to make use of the difference (or ratio) of the spectral absorption characteristic values at II and to minimize changes in the spectral characteristic values due to irregular reflection within the diamond. A method for observing reflected light of a broad wavelength is shown, and it is said that it is possible to identify whether a diamond is naturally derived or artificially derived. Regarding this concept, it was confirmed that the image of the diamond fine abrasive grains reflected together with the base metal, the binder and the filler that we were trying to identify could be selected and identified.

  In other words, from the viewpoint of comparing the reflected light in the multi-frequency wavelength region described in Patent Document 5 and classified as I in this specification by ratio, it is also shown in Patent Document 6 published before that. The wavelength range to be compared is a visible light range including a wavelength of 415 nm in Patent Document 5 and a near-infrared region in Patent Document 6, and for artificial diamonds to be identified, There is a problem that the reflectance is not clearly changed from the visible light region to the infrared light region and cannot be applied.

  On the other hand, since the top surface is in a random position because the diamond fine abrasive grains are not processed, the visible surface reflected in the top surface of the processed diamond described in Patent Document 5 and classified as II in the present specification is originally used. Although it is difficult to take an observation posture of observing reflected light in the light wavelength range, it has the same effect as reflecting light from the diamond top surface directly into the camera. FIG. 5 shows an example when the reflected light passing therethrough is captured and an image / image obtained by the reflected light is acquired. This example is carried out for one diamond fine abrasive grain. However, the reflection intensity on the regular reflection surface exceeds the imaging limit value, but on the surface where there is no reflection, the image is taken under. In addition, there is no distinction between this non-reflective surface and images other than diamond fine abrasive grains that do not reflect fillers, and accurate data regarding the size and shape of the diamond fine abrasive grains cannot be obtained. There is a problem that the image is insufficient in order to convert the cutting edge characteristics into numerical data.

  The present invention has been made paying attention to the above circumstances, and incorporates a method for automatically identifying diamond fine abrasive grains from those obtained in the obtained image of the processing contribution surface of the precision abrasive tool. Another object is to provide a diamond fine abrasive image extraction device.

  A method for acquiring a shape image of diamond fine abrasive grains of a precision abrasive tool comprising diamond fine abrasive grains, a binder, and a filler containing at least carbon, and having an ultraviolet light region having a wavelength of 400 nm or less and a visible light region having a wavelength of 500 nm or more The difference between the UV and visible luminance values for all pixels within the same angle of view is calculated from the image obtained by illuminating each light source with illumination so that the intensity is substantially uniform from all directions. The difference brightness value is used for threshold determination, and it is determined that pixels larger than the threshold are pixels in which diamond fine abrasive grains are copied, and those pixel aggregates are regarded as an image of one diamond fine abrasive grain. An abrasive distribution image is obtained.

  In order to identify the diamond fine abrasive grains, the average value of the luminance of all the pixels constituting the pixel aggregate extracted as diamond fine abrasive grains is defined as the “pixel average value”, and the visible light pixel average value by visible light is further defined. The ratio of the average pixel value of UV light and the average pixel value of ultraviolet light is defined as the “pixel average value ratio”, and the pixel aggregate whose pixel average value ratio is equal to or greater than the threshold value is judged to be truly a copy of the diamond fine abrasive grains. To do.

  Using the diamond fine abrasive image extraction method described in the previous section, the diamond fine abrasive image is extracted based on the image of the processing contribution surface of the precision abrasive tool tip. Get the device.

  Among the pixel aggregates that have been identified as diamond fine abrasive grains by acquiring images when the visible light is applied as incident light to the processing contribution surface of the precision abrasive tool and the diamond fine abrasive grain image extraction device of the previous section A grinding tool grinding | polishing surface inspection system provided with the grindstone cutting edge characteristic observation apparatus which has a function which analyzes the shape of the part with a high luminance value is obtained.

  According to the method of the present invention, an image created by extracting only diamond fine abrasive grains involved in processing from an image obtained by imaging a processing contribution surface of a precision grindstone made of diamond fine abrasive grains, a binder, a filler, and a base metal. By analyzing this image, information regarding the shape of the diamond fine abrasive grains can be provided as digital information.

One configuration example of diamond fine abrasive image extraction device Example of measurement results of spectral characteristics of diamond fine abrasive grains, filler and binder Example of a magnified image of a part of the processing-contributing surface of a resinoid bond diamond wheel chip with a general optical microscope camera Example of an enlarged image of part of the processing-contributing surface of an aligned abrasive tool using a general optical microscope camera Sample image of diamond fine abrasive grains imaged with visible light reflection Flowchart of diamond fine abrasive grain extraction process incorporated in the device of the present invention Example of diamond fine abrasive grains imaged by irradiating with ultraviolet light source Example of diamond fine abrasive grains imaged by irradiating a visible light source Sample image taken with an ultraviolet light source Sample image taken with a visible light source Sample image after difference processing Sample image after binarization Sample image after labeling Sample image after gap interpolation processing Sample image after diamond fine abrasive image extraction processing (A) is a sample image example after the labeling process, (b) is a sample image example obtained by extracting an image from (a) using a pixel average value ratio of 1.75 as a threshold value.

  When the reflectance from the ultraviolet light region to the infrared light region (wavelength 300 nm to 1200 nm) is measured for the diamond fine abrasive grains, the binder, and the filler, the spectral characteristics as shown in FIG. 2 are shown as an example. For the binder and filler, the reflectance is almost uniform in all measured wavelength ranges, but only the diamond fine abrasive grains have different reflectances in the wavelength region of 400 nm or less and the wavelength region of 500 nm or more. . In addition, although the measurement example of the base metal is not listed here, the reflectivity of the iron-based material generally used for the base metal is the same as the binder and filler in the ultraviolet light region to the infrared light region. It is confirmed that it is substantially uniform. While the reflection spectrum of components other than the diamond fine abrasive grains contained in the tool work surface of this precision abrasive tool has little change in a wide wavelength range from the visible light range to the infrared light range, the diamond fine abrasive grains Using the phenomenon that only the reflectance spectrum has a clear inflection point, and following the concept below, the diamond fine grain is based on the pixels where the diamond fine abrasive grains are reflected in the entire image obtained. It is possible to selectively extract an image of abrasive grains.

Next, the concept of identifying a pixel in which diamond fine abrasive grains are reflected will be described.
Depending on the imaging conditions, the absolute value of the reflection intensity from the diamond fine abrasive grains, binder, filler, and base metal itself varies, so the absolute value of the incident luminance value stored in the pixel of the digital image data can be identified. It cannot be handled for this purpose. Therefore, at the same angle of view, two photographs are taken in two wavelength regions where the reflectance characteristics of the diamond fine abrasive grains are different, for example, in the ultraviolet light region (wavelength of about 350 nm) and in the visible light region. The ratio is taken for all pixels at the same position where there is reflection, and it is determined that the diamond fine abrasive grains are reflected in the pixels whose ratio is equal to or greater than a certain value.

  In addition, by simply shining light from one direction on the diamond fine abrasive grains, the reflection intensity on the regular reflection surface may exceed the imaging limit value, but on the surface where there is no reflection, it may be taken under and under the image, Since there is no distinction between this non-reflective surface and images other than diamond fine abrasive grains that are not reflective by fillers, and accurate data on the size and shape of diamond fine abrasive grains cannot be obtained, light irradiation Not enough as a method. In order to solve this problem, it is effective to irradiate light of almost uniform intensity from all directions. As a method for this, for example, a dome illumination or the like that is arranged so that substantially uniform light can be applied to an object to be photographed by appropriately arranging ultraviolet light and visible light with an LED or the like is used. Incorporate into the extractor.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.
FIG. 1 shows a configuration example of the present invention. The diamond fine abrasive grain image extracting device main body 1 is set up as vertically as possible on the upper surface of the device base assembly 10 on the upper surface of the device base assembly 10 storing the power supply device, sequencer control device, and the like in the device. The Z-axis assembly 20 and the X-axis table 51 that can slide with high positional accuracy in the horizontal direction (X-axis direction) on the upper surface of the apparatus base assembly 10 and the observation sample (precision abrasive tool) are positioned on the X-axis table 51 with high accuracy. An X-axis precision slide table on which a rotatable sample rotation table assembly 60 (or a sample table assembly 61 that can slide with high accuracy in a direction orthogonal to the X axis may be used, but will be described here as a configuration example). Personal computer capable of high-capacity and high-speed processing for assembling 50 and acquisition control of captured images by optical camera 31, storage of acquired images, and analysis of acquired images Hereinafter, a keyboard 93 is an input mechanism to the display device 92 and PC 91 for displaying the status of the called "PC".) 91 and PC 91. The device base assembly 10, the PC 91, the display device 92, and the keyboard 93 are connected by dedicated cables. In the present invention, the diamond fine abrasive grain image extracting device main body 1 and the PC 91, the display device 92, and the keyboard 93 may be integrated or separately provided. This is because it is a design as a diamond fine abrasive grain image extraction device and does not affect the functions and performance required here.

  The Z-axis assembly 20 is attached to the support 21, the linear scale 22 that moves along the support 21 in a vertical direction with high accuracy, the Z-axis servo motor 23 that drives the support, and the linear scale 22. The optical dome 31 is composed of an optical camera 31 and a lens 32 capable of imaging up to the ultraviolet light region, and LED light sources for both ultraviolet light and visible light fixed to the column 21. In order to prevent the optical camera 31 and the lens 32 moving up and down from colliding with a sample or the like, it is effective to provide an optical limiter for driving the Z-axis servomotor 23. The attached optical camera 31 is a monochrome photographing camera and can record an image with 255 gradations (8 bits) or more. However, the optical camera may be a color imaging camera, and at that time, it is necessary to perform a gray scale conversion process described later in software.

Although the illumination method used here is not limited to the LED dome illumination 41, it is assumed that the following functions are satisfied.
(1) An ultraviolet light source having a wavelength shorter than 400 nm (for example, a 365 nm LED or a xenon lamp that transmits ultraviolet light through a bandpass filter) is incorporated. (2) A visible light source (for example, a wavelength of 525 nm (green) ) Incorporate a xenon lamp that transmits visible light through an LED or a bandpass filter. (3) For ultraviolet light and visible light illumination, position the irradiation surface (observation) by arranging a number of light source elements alternately. The structure is such that the brightness distribution on the surface of the sample does not occur as much as possible. Furthermore, the uniformity of the amount of light reaching the irradiation surface (illuminance) is secured by using / adding a diffusion dome or the like.
(4) Irradiation switching between ultraviolet light and visible light can be performed by an electric switch or the like so that the position of illumination does not have to be changed by imaging.
(5) It is assumed that the irradiation light quantity can be adjusted digitally so that the camera image sensor has a luminance of 0 to 255 gradations.

  In the combination of the optical camera 31 and the lens 32 capable of imaging up to the ultraviolet light region, in a camera system that is usually commercially available, the focal position by ultraviolet light and the focal position by visible light are slightly different. In order to correct this, it is necessary to reduce or enlarge one image by digital processing, or to adjust the field angle range to be imaged to the same by a lens magnification changing function. To further refine and automate this, a special design of the lens system, or a specially designed camera such as two parallel lens systems for ultraviolet light and visible light and one image sensor To do.

  In the slide table assembly 50, a sample rotation table assembly 60 composed of a rotary precision drive motor is placed on a slide table in which an X-axis precision slide table 51 and an X-axis servo motor 52 are combined, and an observation sample is accurately measured. An angular position can be set. As the accuracy of the position and angle position setting, images are acquired with two wavelengths in the visible light region and the ultraviolet light region at the same angle of view, and each pixel position of each image is a two-dimensional coordinate position shift on the acquired image. Is 1 pixel or less.

  The observation sample is moved to the X-axis position, which is the imaging position, by the high-precision X-axis positioning function of the X-axis precision slide table 51, and the visible light source of the LED dome illumination 41 is lit and irradiated. In this state, the optical camera 31 is focused by the Z-axis servomotor 23. When focus is achieved, the Z-axis operation is stopped, a visible light image is acquired, an image name is assigned, and the image is stored in the PC 91. Subsequently, the LED dome illumination 41 is switched to the ultraviolet light source. At this time, in the commercially available camera, the focal position of the optical camera 31 is slightly shifted between visible light and ultraviolet light. Therefore, when the focus of the optical camera 31 is finely adjusted by the Z-axis servomotor 23, the operation of the Z-axis is stopped and the ultraviolet light image is displayed. Is obtained, and an image name is assigned and stored in the PC 91, whereby two images are obtained. The angle-of-view size correction process may be performed before the image is stored, or may be additionally performed during the diamond identification process in the subsequent process.

  Next, the sample rotation table assembly is rotated by one angle of view, and the observation sample is moved to the X-axis position as the imaging position by the high-precision X-axis positioning function of the X-axis precision slide table 51, and the LED dome illumination 41 is visible. With the light source turned on and illuminated, the optical camera 31 is focused, a visible light image is acquired, an image name is assigned, and the image is stored in the PC 91. Subsequently, the LED dome illumination 41 is switched to the ultraviolet light source, and when the focus shift between visible light and ultraviolet light is finely adjusted, an ultraviolet light image is acquired, and an image name is given and stored in the PC 91 repeatedly. Obtain two images with the angle of view.

  Further, the operation of rotating the sample rotation table to acquire visible light and ultraviolet light images is continued N times until the observation sample rotates once, thereby obtaining N × 2 images.

  Subsequently, the precision slide table 51 is moved by one angle of view of the optical camera 23 by the X-axis servo motor 52 of the X-axis slide table 50 in the radial direction of the processing-contributing surface of the observation sample, and the entire processing-contributing surface width is obtained. The imaging is repeated until an image is obtained.

  All the images picked up as described above are respectively combined with the visible light image and the ultraviolet light image to obtain two images of the entire area of the processing contribution surface of the observation sample, and the process proceeds to the next image analysis process.

  It should be noted that in the image obtained in this manner, the diamond fine abrasive grains are different in luminance due to the difference in imaging wavelength shown in FIG. 7 for ultraviolet light and FIG. 8 for visible light. It is reflected in the state of. However, up to this point, the filler is also reflected in the same luminance state (black state) in the ultraviolet light. Hereinafter, a method for identifying diamond fine abrasive grains from images picked up in this way will be described using image examples obtained by applying the method of the present invention.

FIG. 6 shows an example of a diamond extraction processing software flow for extracting diamond fine abrasive grains incorporated in the PC 91 according to the present invention. Based on this, the concept of the method for realizing each processing program will be described. Of course, it is not limited to the contents of the flow and the program processing, and an image captured with two wavelengths including the ultraviolet light wavelength region (or an image with a larger number of wavelengths including the ultraviolet light wavelength region may be used). It may be realized that an image of diamond fine abrasive grains is extracted from the image.
9 to 16, for the sake of explanation, diamond fine abrasive grains are appropriately arranged on a metal plate that simulates a base metal, and the results of analysis by the method applied to the present invention are shown for each procedure. It is shown as an example of an obtained image. It is not always necessary to store all these images in the PC 91.

  By using the apparatus of the present invention, an image (FIG. 9) in a state irradiated with ultraviolet light and an image (FIG. 10) in a state irradiated with light in the visible light region are acquired. This image is assumed to be captured by a monochrome camera in this example, but it may be captured by a color camera. If these acquired images are captured by a color camera, a “grayscale process” is performed to perform monochrome conversion. Process into an image.

  In the following description, the ultraviolet light irradiation image (FIG. 9) after gray scale conversion will be referred to as “image A”, and the visible light irradiation image (FIG. 10) will be referred to as “image B”. This designation is merely for explanation, and the designation of the following images is the same. In the image B, those that appear light white and look like constellation stars are diamond fine particle candidates, and these appear as black spots at the same position in the image A because most of the abrasive grains are less reflective. Some of them have relatively strong reflection and light white color. In addition, in image A and image B, in addition to the diamond fine abrasive grain candidates, several streaks on the surface of the base metal are shown in white lines.

  Next, as “saturation adjustment processing”, an average saturation value of each base metal part is obtained so that the brightness of the base metal parts of the images A and B is approximately the same, and all of the images A The average value is added to the pixel luminance value. Alternatively, when the image B is brighter, all pixel luminance values of the image A are reduced. This is an unnecessary process if the intensity of the illumination is adjusted so that the brightness at the standard position (desirably the base metal surface is desirable) in the images A and B at the time of imaging.

Next, as “difference processing”, the luminance values of the pixels (pixels) at the same position in the images A and B are acquired,
Luminance value difference = | Luminance value of image B−Luminance value of image A |
Calculate This difference processing is performed for all pixels. A gray image (FIG. 11) with the difference value thus obtained is obtained. This image is referred to as “image C”.

  Next, as a “binarization process”, for each pixel of the image C, for example, a binary conversion is performed in which a pixel value having a luminance difference value greater than 30 is a white pixel, and otherwise a pixel value is a black pixel. A black and white binary image shown in FIG. 12 is obtained. This is referred to as “image D”. In the case of this image example (image D), the scratches on the base metal remain and appear. Note that the threshold value of the brightness difference value is determined in advance by the type of precision abrasive tool.

  Next, as the “labeling process”, the same number is assigned to the pixels gathered among the white pixels of the image D (hereinafter referred to as “labeling”). Count the number of constituent pixels of a pixel assembly with the same labeling, and set the threshold for the number of pixels that make up the diamond fine abrasive grains of average size from the average grain area of the diamond fine abrasive grains and the element size of the pixels. Then, only the label within the threshold range is left and re-imaged. (FIG. 13). This image is set as an image E. By passing through this processing, labeling with a small number of constituent pixels (= noise components such as linear flaws) is eliminated. However, in the image E, on the contrary, it is erroneously determined as noise, and a part (black) missing in the white particle area is generated.

  Next, “gap interpolation processing” is performed to supplement the above-described missing portion. The black portion taken into the interior surrounded by the labeling process is assumed to have the same property as the surrounding area labeled, and given the same number as the surrounding label (labeling) and re-imaged, FIG. 14 is obtained. This is image F. However, the white portion of the image F may still contain things other than diamond fine abrasive grains.

Next, the following concept of “pixel average value ratio” is introduced as a threshold value for diamond fine abrasive grains.
The pixel group of diamond fine abrasive grain candidates that have been labeled in the previous procedure includes pixels that have been captured and recorded with the luminance value saturated on the over side or under side, so they are in the same position. Some pixels have values that have no physical meaning by taking the ratio between pixels. In order to avoid this, the luminance is averaged as fine abrasive grains composed of the same labeled pixel aggregate and defined as “pixel average value”, and the pixel average values are compared. Further, the ratio of the pixel average value based on the visible light image and the pixel average value based on the ultraviolet light image is defined as “pixel average value ratio”.
Using this pixel average value ratio, the following “diamond image extraction process” is performed.

  In the diamond fine abrasive grain image extraction process, this threshold judgment is applied to process the image F to obtain FIG. This is called an image G. The remaining particles in the image G are identified as diamond fine abrasive grains.

  FIG. 16 shows an example in which the threshold value of the pixel average value ratio is examined. FIG. 16A is an image of the processing contribution surface that has been subjected to the labeling process, and FIG. 16B is a gap interpolation process performed from the image of FIG. 16A with a pixel average value ratio of 1.75 as a threshold value. This is an example of image acquisition. In the image of (a), the slightly upper particle on the left of the screen is dropped in (b), and the pixel average value ratio of 1.75 is too large as a threshold value. Next, when the pixel average value ratio was 1.4, extraction was possible with a sufficient recognition rate. In the case of this sample, it was found that it is appropriate to determine the diamond fine abrasive grains with a threshold value of 1.4. As described above, this threshold value needs to be set by confirming validity in advance by the type of precision abrasive tool. For example, a value of about 1.3 to 1.6 should be adopted. Is desirable.

  In addition, according to the diamond fine abrasive image extraction technology described so far, when the abrasive grain image is smaller than the labeling determination threshold because the projection amount of the diamond fine abrasive grain is assumed in the actual processing contribution surface, Although there is a possibility that it is not a diamond fine abrasive grain, it may be said that it is an abrasive grain having a small degree of involvement in processing in the first place, and is within an allowable extraction error range.

  In addition to these, in order to highlight the cutting edge shape involved in the processing of diamond fine particles in the captured image, an image obtained by applying visible light incident on the processing-contributing surface was obtained, and the pixel was determined to be a diamond fine abrasive grain If the shape of the directly facing part with respect to the camera lens having a high luminance value inside the aggregate is analyzed, a grinding surface inspection system that analyzes the performance of the abrasive cutting edge involved in processing with high accuracy can be obtained.

DESCRIPTION OF SYMBOLS 1 ... Diamond fine abrasive image extraction apparatus main body 10 ... Apparatus base assembly 20 ... Z-axis assembly 21 ... Strut 22 ... Linear scale 23 ... Z-axis servo motor 31 ... Optical camera 32 ... Lens 41 ... LED dome illumination 50 ... Slide table Assembly 51 ... X-axis precision slide table 52 ... X-axis servo motor 60 ... Sample rotation table assembly 61 ... Y-axis precision slide table assembly 62 ... Sample table 91 ... Personal computer 92 ... Display device 93 ... Keyboard

Claims (4)

  A method for acquiring a shape image of diamond fine abrasive grains of a precision abrasive tool comprising diamond fine abrasive grains, a binder, and a filler containing at least carbon, and having an ultraviolet light region having a wavelength of 400 nm or less and a visible light region having a wavelength of 500 nm or more The difference between the UV and visible luminance values for all pixels within the same angle of view is calculated from the image obtained by illuminating each light source with illumination so that the intensity is substantially uniform from all directions. The difference brightness value is used for threshold determination, and it is determined that pixels larger than the threshold are pixels in which diamond fine abrasive grains are copied, and those pixel aggregates are regarded as an image of one diamond fine abrasive grain. A method for obtaining a distribution image of abrasive grains.   The method according to claim 1, further comprising: defining an average value of luminance of all pixels constituting a pixel aggregate extracted as diamond fine abrasive grains as a "pixel average value", and a visible light pixel average value by visible light; The ratio of the average pixel value of ultraviolet light by ultraviolet light is defined as “pixel average value ratio”, and a pixel aggregate whose pixel average value ratio is equal to or greater than a threshold value is judged to be truly a copy of diamond fine abrasive grains. A method for obtaining a distribution image of diamond fine abrasive grains.   An apparatus using the method of claim 1 and claim 2, wherein an image of a diamond fine abrasive image for extracting an image of a diamond fine abrasive grain based on an image obtained by imaging a processing contribution surface of a chip of a precision abrasive tool. Extraction device.   An apparatus for extracting a diamond fine abrasive image according to claim 3 and an image obtained by irradiating visible light as incident light on a processing-contributing surface of a precision abrasive tool. A grinding tool grinding surface inspection system comprising a grinding wheel cutting edge characteristic observation device having a function of analyzing the shape of a portion having a high luminance value.

Images