JP5536979B2 - Roll mill roll cleaning endpoint inspection equipment - Google Patents

Description

The present invention relates to a roll mill used to knead and disperse paste-like processing materials in the manufacturing process of electronic parts, printing inks, paints, ceramics, chemicals, foods, and various other products. The present invention relates to an end point inspection apparatus for roll cleaning of a roll mill that can inspect the surface of the roll mill.

  Conventionally, in the manufacture of products as described above, roll mills have good dispersibility and are easy to clean, change color and change varieties because they are open, and are frequently used in the manufacture of small quantities of many types of products. Has been. Such a roll mill has a plurality of rolls, for example, a three-roll mill (see, for example, Patent Document 1), each having a rear roll, a middle roll, and a front roll having different peripheral speeds, and a pasty roll between adjacent rear rolls and middle rolls. It is configured to supply the processing material, disperse the processing material by the compression and shear action generated between the rolls, send it to the front roll, and take it out from the front roll with a doctor knife, etc. doing. In this cleaning operation, the cleaning cloth such as a cleaning solvent is soaked in the waste cloth, wipe the cloth surface manually with the waste cloth, and wipe it again and again. Had ended. Usually, a metal roll such as a chilled roll is used as the roll. However, in the case of manufacturing a product that does not require metal contamination (contamination), a ceramic roll having a surface made of a ceramic material is used. In the case of such a ceramic roll, after visually determining the end point of cleaning, a metal spot spray further containing ammonium thioglycolate, nonionic surfactant, cleaning aid, fragrance, etc. is applied to the roll surface. The cleaning result was checked by spraying to confirm the discoloration, but often a cleaning residue was seen. As described above, since the roll mill was manually cleaned and the cleaning endpoint was visually determined, human error and unevenness may occur, or contamination may be brought into the next process. It was.

  Also, roll mill rolls compress and shear processing materials between rolls, which may cause roll surface roughness to increase and scratches and patterns to occur, so it is necessary to replace or overhaul in a timely manner. Conventionally, the time of such work is also judged visually, so there are judgment irregularities and errors.

Furthermore, since the processing material includes various materials having different colors, brightness, and saturation, it may be difficult to easily determine the state of cleaning or the like due to the relationship with the roll surface.
JP 2006-305435 (FIG. 2)

The problem to be solved by the present invention is that of a roll mill that can accurately determine the cleaning end point by inspecting the roll surface even if the roll material is a chilled roll, a ceramic roll or the like in a roll mill such as a three-roll mill. It is to provide an end point inspection apparatus for roll cleaning .

According to the present invention, a light source that irradiates the roll surface of a roll mill, a camera that shoots the roll surface, and a hue distribution image is checked based on the image captured by the camera to detect dirt, roughness, and scratches on the roll surface. Provided are an image processing apparatus and an end inspection apparatus for roll cleaning of a roll mill characterized by inspecting an end point of roll cleaning based on a state of a roll surface detected by the image processing apparatus. .

According to the present invention, the image processing apparatus, in the case of slightly different processing materials and color saturation is low in achromatic processing material with transparency but the roll surface has a color, a hue distribution image endpoint inspection apparatus roll cleaning of the roll mill so as to identify the residual presence and residual places Ji Ekku to process material is provided in a specular distribution image, the problem can be solved.

The present invention is configured as described above, a light source that irradiates the roll surface of a roll mill, a camera that photographs the roll surface, and a hue distribution image that is checked by an image captured by the camera to check for contamination and roughness of the roll surface. Since the image processing apparatus for detecting scratches and the roll cleaning end point are inspected based on the state of the roll surface detected by the image processing apparatus, the human error is eliminated and the contamination is eliminated in the roll mill cleaning process. The endpoint can be detected by objective evaluation, which can make a decision to proceed to the next process or the next work without causing unevenness or waste, improving the quality and workability of the cleaning, The end of the roll can also be accurately determined.

  If the hue of the processing material and the roll surface is different, if the hue distribution image is checked by the image processing device, the progress of cleaning will not be affected by the uneven illumination that irradiates the roll surface. By displaying the color or the hue range of the processing material, the remaining location can be easily identified by changing the hue of the processing material from the hue of the processing material to the surface of the roll, and cleaning is performed from the ratio of the number of pixels that fall within the hue range of the processing material. You can see the degree of progress.

  If the hue of the processing material and the roll surface is similar, illuminate with a color filter or perform image processing similar to that using a color filter by image processing and dye the roll surface with the color of the color filter. When the hue distribution image is checked by the image processing apparatus, the hue of the processing material changes from the hue of the processing material to the hue of the roll surface colored by the filter effect as the cleaning progresses, and the state is changed to the color or the processing material. By displaying the hue range, the remaining location can be easily identified, and the progress of the cleaning can be understood from the histogram of the hue range of the processing material.

  In the case of an achromatic processing material with transparency, when checking the hue distribution image and the specular reflection distribution image, the image of the processing material has a low saturation but a hue, and the roll surface and the hue are slightly Since they are different, a change in hue distribution is obtained. Further, the specular reflection component present in the lightness distribution becomes elliptical reflecting the shape of the roll due to the damping effect of the treatment material. During cleaning, the shape changes irregularly to reflect cleaning unevenness. In the roll area where the hue has changed, it is possible to determine the presence or absence of washing by evaluating the magnitude of hue fluctuation and the change in brightness, and to detect the progress of washing from the brightness distribution of specular reflection. In addition, in the case of an achromatic processing material without transparency, if the hue distribution image and the lightness distribution image are checked, the lightness increases in the roll area where the processing material is removed as the cleaning progresses. Causes hue or becomes clear. A change in brightness distribution is detected by a difference image or the like, and a change in hue distribution is detected by an average value or distribution of hues, whereby the progress of cleaning and the remaining location of the processing material can be specified.

Figure 1 shows an embodiment of the present invention, inspection apparatus of the present invention includes a light source 2 for illuminating the roll surface of the roll mill 1, a camera 3 for photographing the roll surface, the captured image of the camera It has an image processing device 4 such as a notebook personal computer that detects dirt, roughness, scratches, etc. on the roll surface. Based on the state of the roll surface detected by the image processing device, the end point of the roll cleaning and the end of the roll are determined. judge.

  In the figure, the roll mill is shown as a three-roll mill 1 having a rear roll 5, a middle roll 6, and a front roll 7. The light source 2 uses indoor lighting such as a white fluorescent lamp and a strobe of the camera 3 in combination. However, in the case of indoor lighting in which automatic white balance adjustment is not sufficiently effective, an independent strobe light source or an illumination light source with a known color temperature can be used.

  As the camera 3, a digital camera having an exposure amount and white balance adjustment function is used, and in particular, a digital camera having a color specification independent of the model is preferable. It is also possible to use a high-resolution video camera in place of the digital camera, and it is desirable to be able to operate setting of the shooting conditions of the camera, shooting, transfer of image data, etc. from the personal computer. These controls and image processing may be performed using existing software or a general-purpose program language.

  As shown in FIG. 2, it is preferable to photograph the surface of the roll centering on the center of the middle roll from the center of the three rolls as shown in FIG. 2, but for the sake of convenience, the shadow of the roll mill frame is taken. In order to prevent this, the image is taken from the front of the roll mill or from the opposite direction, and as shown in FIG. 3, the image is taken within a range of about 45 ° before and after the three rolls are sufficiently visible in the field of view.

  When photographing a roll mill with a large specular reflection by illumination of the light source 2, a plurality of light sources are required. For example, a strobe attached to the camera and room lighting may be used in combination, or the strobe may be simultaneously emitted from a plurality of directions, or may be photographed using a plurality of illumination light sources. In the above shooting, the entire roll is shot several times while rotating the three rolls. If the white balance of multiple light sources is different, adjust using a color temperature conversion filter.

  Thus, when the hues of the roll surface and the processing material are different, the roll mill being cleaned is photographed with a digital camera under illumination of a white light source, and the color image is sent to an image processing apparatus (personal computer). The RGB image is converted into an HSV image above, and the hue (H) distribution image is compared with the hue distribution image of the initial roll surface to which the processing material is not adhered, and the washing hue from the changing hue image or hue gamut image. It is possible to clearly know whether or not the processing material remains and where it remains.

  When the hue of the roll surface and the processing material is similar, an optical color filter is installed in front of the white light source or the same processing as that using the optical color filter is performed in the image processing apparatus, and the roll surface is subjected to the color filter. Dyed in the color. The roll surface is dyed in the color of the color filter when the image saturation is low, but the processing material is usually not affected by the high image saturation, and the hue does not change. As a result, there is a hue difference between the portion where the processing material remains and the roll surface, so that the processing material adheres to the hue (H) distribution image obtained by the camera in substantially the same procedure as described above. Compared with the hue distribution image of the initial roll surface that is not present, it is possible to accurately know whether or not the processing material remains during cleaning and the remaining location from the changing hue image or hue area image.

  When the processing material is an achromatic color system with transparency, the degree of progress of cleaning and the remaining place are specified from changes in the hue distribution image and the specular reflection distribution image. That is, even in the case of an achromatic processing material, the chroma has a low color but has a color, and the color often differs slightly from the color of the roll surface. In the image of the processing material attached to the roll surface, the hue changes smoothly and smoothly around a predetermined value, while the hue of the image on the roll surface often varies randomly. In addition, due to the difference in reflectance between the treatment material and the roll surface, the brightness of the image often differs depending on whether or not cleaning is performed. Therefore, in such a case, it is possible to identify the cleaning region by detecting the region where the hue has changed from before cleaning in the cleaning, and evaluating the change in hue and the change in brightness there. In the case of a treatment material having transparency, the component of specular reflection that occurs when the transmitted light is reflected by the roll surface becomes significant. This component reflects the shape of the roll due to the damping effect according to the thickness of the treatment material, and becomes an ellipse that is long in the axial direction of the roll. On the other hand, the specular reflection component of the roll itself has a linear shape extending in the axial direction. In such a case, the progress of cleaning can be detected from the ratio of the length in the circumferential direction and the axial direction of the pattern in which the specular reflection component is binarized. The pattern being cleaned changes irregularly reflecting uneven cleaning, and the degree of progress can be detected from the change.

  In the case of an achromatic treatment material with low transparency of the treatment material, the progress of cleaning and the remaining location of the treatment material can be detected from the change in the hue distribution image and the brightness distribution image on the roll surface. In the case of such a processing material, the brightness of the image is usually increased in the roll area where the processing material has been cleaned, where a hue occurs or becomes clear. The change in brightness distribution is detected by forming a difference image from the image before washing, and the change in hue distribution is obtained by calculating the average value and distribution of hue in the area where the brightness change has occurred, and the previous past of the roll Detect by comparing with data.

  In order to determine the roll overhaul time, various unique phenomena at the end of the roll are stored in an image database, and each image can be determined by referring to this database. In other words, the roll near the overhaul has a rough surface, no specular reflection, scratches in the circumferential direction, etc., and a unique pattern. To create a database. In order to determine overhaul, first, an image of each roll surface is sequentially processed for roughness, scratches, and patterns, collated with a database, and a diagnostic result is output.

It is explanatory drawing which shows one Example of the apparatus of this invention. It is explanatory drawing seen from the plane of the roll mill. It is explanatory drawing seen from the side of the roll mill.

Explanation of symbols

1 Roll Mill 2 Light Source 3 Camera 4 Image Processing Device 5 Rear Roll 6 Middle Roll 7 Front Roll

Claims (1)

A light source that irradiates the roll surface of a roll mill, a camera that images the roll surface, an image processing device that checks a hue distribution image based on an image captured by the camera and detects dirt, roughness, and scratches on the roll surface; An end point inspection device for roll cleaning of a roll mill that inspects an end point of roll cleaning based on the state of the roll surface detected by the image processing device, wherein the processing material attached to the roll surface is transparent. If it Ru Oh achromatic system is different treatment material and color of the roll surface has a color, the roll surface treatment material is detected hue distribution image and the specular reflection distribution image of the initial roll surface unattached The end point of roll cleaning of a roll mill characterized by identifying the presence / absence and the location of the treatment material from the changes in the hue distribution image and specular reflection distribution image Door inspection apparatus.

Images