Ceramic body defect detection device and defect detection method
Patent Information
- Authority / Receiving Office
- DE · DE
- Patent Type
- Patents
- Current Assignee / Owner
- NGK INSULATORS LTD
- Filing Date
- 2017-11-13
- Publication Date
- 2026-07-09
Abstract
Description
technical field
[0001] The present invention relates to an apparatus and a method which inspect for the presence of a defect on an outer surface of a ceramic body, and particularly relates to an apparatus and a method which inspect an end face of a honeycomb structural body. State of the art
[0002] A honeycomb structural body, which is a ceramic porous body (ceramic body), is widely used as a filter that captures particulate matter contained in exhaust gas from an internal combustion engine, a boiler system or the like, or a catalyst carrier of an exhaust gas purifying catalyst. The honeycomb structural body includes a plurality of honeycombs partitioned by partition walls respectively extending along the axial direction of the structural body and surrounded by a tubular outer surface (outer wall). The ceramic honeycomb structural body is excellent in heat resistance, thermal shock resistance and oxidation resistance, and hence is widely used for the above-mentioned applications and the like.
[0003] In some honeycomb structural bodies, honeycomb openings on both end faces are alternately closed (in a checkerboard pattern) (also referred to as plugging) (closed honeycomb structural bodies). Such a sealed honeycomb structural body is used for a DPF (diesel particulate filter), for example.
[0004] A ceramic honeycomb structural body is typically formed by extrusion molding a clay-like raw earth obtained by kneading a powder of ceramics (e.g. cordierite, SiC or alumina) which is a base material therein with, e.g., an organic binder and water, and firing a honeycomb shaped body thus obtained manufactured. When a closure is to be added, it can be done, for example, in a manner that an end part of a honeycomb fired body on which a honeycomb not to be closed has been masked in advance is immersed in a slurry-like filler to close an opened honeycomb with the filler and then the honeycomb fired body is fired again (see Patent Document 1, for example). Alternatively, an unsealed honeycomb formed body is filled with the filler as described above and then fired to thereafter obtain a sealed honeycomb structural body.
[0005] A honeycomb structural body manufactured by such a method is inspected to ensure that there is no defect such as a crack, a chip or a hole at a side face, an end face with an opening and an inner partition wall, and then shipped as a product.
[0006] Patent Document 1 discloses, as a method for inspecting a sealed portion of a sealed honeycomb structural body, a method of imaging a honeycomb from one face while light is irradiated thereon on the other face, and detecting a defect at the sealed portion based on the gray scale (luminance) of light obtained by performing image processing on a captured image thus obtained.
[0007] In another publicly known method (see, for example, Patent Document 2), a telecentric optical system and a camera whose optical axis is aligned with that of the optical system are on the one end part side of the honeycomb structural body in one with respect to the axis line direction of the honeycomb structural body a predetermined angle in a slanting direction, and the gray level of an image formed by light incident obliquely on a partition wall is identified to perform crack detection on the partition wall.
[0008] When the defect inspection described above is performed using the gray scale appearing in a captured image on an end face of a honeycomb structural body, a defect such as a crack, chipping or hole occurring at the edge of a honeycomb opening must be reliably distinguished from the honeycomb opening itself . In addition, in a case of a sealed honeycomb structural body, normal irregularities (irregularities which do not cause a problem in terms of product specifications) present particularly at a sealed part and a rib part must not be mistakenly recognized as a crack, a chip, a hole and the like.
[0009] For example, it is known that when oblique illumination as disclosed in Patent Document 2 is used in a defect inspection, a defective part such as a chip or a hole is likely to be a dark part (shadow), but a shadow is also likely to appear in a normal irregular one part, and consequently, in the case of defect detection based on the presence of the dark part, the normal irregular part is erroneously recognized as a defect with a high possibility. Publication IndexPatent Documents Patent Document 1: Japanese Patent Application Laid-Open No. 2010-249798 Patent Document 2: Japanese Patent Application Laid-Open No. 2008-139052 short description
[0010] The present invention was made in view of the above problems, and an object of the present invention is to provide an inspection method and an inspection apparatus which reliably detect a defect on an outer surface of a ceramic body such as an end face of a honeycomb structural body and falsely detect a normal surface irregularity as a defect reliably prevent.
[0011] In order to solve the above-described problem, according to a first aspect of the present invention, an apparatus configured to inspect for the presence of a defect on an outer surface of a ceramic body includes: a table on which a ceramic body as an inspection target is to be placed; an image pickup part configured to pick up an image of an inspection target area as at least a part of an inspection target surface of the ceramic body placed on the table in a normal direction of the inspection target area; one or more illumination parts including four or more illumination elements configured to obliquely irradiate the examination target area with illumination light at an identical irradiation angle in respective irradiation directions different from each other and equiangularly spaced from each other around the imaging part; an evaluation image generation part configured to generate evaluation image data for determining the presence of a defect in the inspection target area based on captured image data captured by the image pickup part; and a defect determination part configured to determine the presence of a defect in the inspection target area based on the determination image data. The plurality of lighting elements are switched on and off one after the other. The image pickup part generates a plurality of pieces of captured image data by capturing an image of the inspection target area when each of the plurality of lighting elements is turned on. The evaluation image generating part: generates minimum luminance image data in which a minimum value of luminance values of the plurality of pieces of captured image data is set at an identical pixel position as a luminance value at the pixel position, and then generates the evaluation image data based on the minimum luminance image data; or generates maximum luminance image data in which a maximum value of luminance values of the plurality of pieces of captured image data at an identical pixel position is set as a luminance value at the pixel position, and generates the determination image data based on the maximum luminance image data.
[0012] According to a second aspect of the present invention, in the defect inspection apparatus according to the first aspect, the determination image generation part further includes a luminance correction processing part configured to correct the luminance values of the plurality of pieces of captured image data so that the luminance values at a previous in each defined from the plurality of pieces of captured image data when the plurality of pieces of captured image data are the same, and generates the minimum luminance image data or the maximum luminance image data based on the plurality of pieces of captured image data whose luminance values are corrected by the luminance correction processing portion .
[0013] According to a third aspect of the present invention, in the defect inspection apparatus according to the first or second aspect, the one or more lighting parts includes a first lighting part including, as the plurality of lighting elements, a plurality of first lighting elements each configured to irradiating the examination target area with illumination light at a first irradiation angle of 30° to 60° inclusive, and a second illumination part including, as the plurality of illumination elements, a plurality of second illumination elements each configured to illuminate the examination target area with illumination light at a second irradiation angle from 60° to 85° inclusive.
[0014] According to a fourth aspect of the present invention, in the defect inspection apparatus according to the third aspect, the imaging part generates a plurality of pieces of first pickup image data by performing imaging of the inspection target area for each of the plurality of first lighting elements when each of the plurality of first lighting elements is turned on , and generates a plurality of pieces of second captured image data by performing image capture of the inspection target area for each of the plurality of second illumination elements, when each of the plurality of second illumination elements is turned on, the plurality of pieces are first captured image data and the plurality of pieces second captured image data the plurality of pieces of captured image data different from each other, the determination image generating part generates in the case that the determination image data based on the minimum luminance image data are to be generated, first minimum luminance image data in which a minimum value of luminance values of the plurality of pieces of first captured image data is set at an identical pixel position as a luminance value at the pixel position, and second minimum luminance image data in which a minimum value of luminance values of the plurality of pieces of second capture image data at an identical pixel position as a luminance value at the pixel position is set as the minimum luminance image data and generates the evaluation image data by generating first evaluation image data based on the first minimum luminance image data and generating second evaluation image data based on the second minimum luminance image data; and in the case that the determination image data is to be generated based on the maximum luminance image data, the determination image generating part generates first maximum luminance image data in which a maximum value of luminance values of the plurality of pieces of first captured image data at an identical pixel position as a luminance value at the pixel position is set, and second maximum luminance image data in which a maximum value of luminance values of the plurality of pieces of second captured image data at an identical pixel position is set as a luminance value at the pixel position as the maximum luminance image data and generates the evaluation image data by generating first evaluation image data based on the first maximum luminance image data and generating second determination image data based on the second maximum luminance image data, and the defect determination part determines the presence of a defect in the sub detection target area based on both the first detection image data and the second detection image data.
[0015] According to a fifth aspect of the present invention, in the defect inspection apparatus according to the third aspect, the number of the plurality of first lighting elements is equal to the number of the plurality of second lighting elements, and the first and second lighting elements are arranged such that an irradiation direction of each of the first lighting elements is included vertical plane always contains an irradiation direction of one of the second illumination elements, the plurality of first illumination elements emit illumination light belonging to a first wavelength band, the plurality of second illumination elements emit illumination light belonging to a second wavelength band different from the first wavelength band, each pair of the first and of the second lighting elements having irradiation directions contained in an identical vertical plane are turned on and off simultaneously, the image pickup part generates the Vi el number of pieces of captured image data by capturing an image of the inspection target area when each pair of the first and second lighting elements is turned on, the defect inspection apparatus further includes a separation image generation part configured to perform color separation of each of the plurality of pieces of captured image data to generate a plurality of pieces of first separation image data and a plurality of pieces of second separation image data based on the first wavelength band and the second wavelength band, in the case that the determination image data is to be generated based on the minimum luminance image data, the determination image generation part generates first Minimum luminance image data in which a minimum value of luminance values of the plurality of pieces of first separation image data at an identical pixel position is set as a luminance value at the pixel position and second minimum luminance image data in which a minimum value of luminance values of the plurality of pieces of second separation image data at an identical pixel position is set as a luminance value at the pixel position as the minimum luminance image data, and generates the evaluation image data by generating first evaluation image data on the basis the first minimum luminance image data and generating second determination image data based on the second minimum luminance image data; and in the case that the determination image data is to be generated based on the maximum luminance image data, the determination image generating part generates first maximum luminance image data in which a maximum value of luminance values of the plurality of pieces of first separation image data at an identical pixel position as a luminance value at the pixel position is set, and second maximum luminance image data in which a maximum value of luminance values of the plurality of pieces of second separation image data at an identical pixel position is set as a luminance value at the pixel position as the maximum luminance image data and generates the evaluation image data by generating first evaluation image data based on the first maximum luminance image data and generating second determination image data based on the second maximum luminance image data and the defect determination part determines the presence of a defect in the Inspection target area based on both the first determination image data and the second determination image data.
[0016] According to a sixth aspect of the present invention, in the defect inspection apparatus according to the first or second aspect, the one or more lighting parts includes: a small-angle lighting part including, as the plurality of lighting elements, a plurality of small-angle lighting elements, each configured to irradiate the inspection target area with illumination light at an irradiation angle of 5° to 30° inclusive; a middle-angle illumination part including, as the plurality of illumination elements, a plurality of middle-angle illumination elements each configured to irradiate the inspection target area with illumination light at an irradiation angle of 30° to 60° inclusive; and a large-angle illumination part including, as the plurality of illumination elements, a plurality of large-angle illumination elements each configured to irradiate the examination target area with illumination light at an irradiation angle of 60° to 85° inclusive.
[0017] According to a seventh aspect of the present invention, in the defect inspection apparatus according to the sixth aspect, the imaging part generates a plurality of pieces of small-angle captured image data by performing inspection-target area imaging for each of the plurality of small-angle illumination elements when each is off of the plurality of small-angle lighting elements is turned on, it generates a plurality of pieces of middle-angle pickup image data by performing imaging of the inspection target area for each of the plurality of middle-angle lighting elements when each of the plurality of middle -angle illumination elements is turned on, and generates a plurality of pieces of large-angle pickup image data by performing image pickup of the examination target area for each of the plurality of large-angle illumination elements when each of the plurality of large-angle Bele When the directional elements are turned on, the plurality of pieces of small-angle captured image data, the plurality of pieces of medium-angle captured image data, the plurality of pieces of large-angle captured image data are the plurality of pieces of captured image data different from each other, the determination image generating part generates in the case that the determination image data is to be generated based on the minimum luminance image data, small-angle minimum luminance image data in which a minimum value of luminance values of the plurality of pieces of small-angle captured image data at an identical pixel position as a luminance value of the pixel position, medium-angle minimum luminance image data in which a minimum value of luminance values of the plurality of pieces of medium-angle captured image data is set at an identical pixel position as a luminance value at the pixel position, and large-angle minimum luminance t-density image data in which a minimum value of luminance values of the plurality of pieces of large-angle captured image data at an identical pixel position is set as a luminance value at the pixel position as the minimum luminance image data, and generates the determination image data by generating small-angle detection image data based on the small-angle minimum luminance image data, generating medium-angle detection image data based on the medium-angle minimum luminance image data, and generating large-angle detection image data based on the large-angle minimum luminance image data ; and in the case that the determination image data is to be generated based on the maximum luminance image data, the determination image generating part generates small-angle maximum luminance image data in which a maximum value of luminance values of the plurality of pieces of small-angle captured image data at a identical pixel position is set as a luminance value at the pixel position, mean-angle maximum luminance image data in which a maximum value of luminance values of the plurality of pieces of mean-angle captured image data is set at an identical pixel position as a luminance value at the pixel position, and larger -angle maximum luminance image data in which a maximum value of luminance values of the plurality of pieces of large-angle captured image data at an identical pixel position is set as a luminance value at the pixel position as the maximum luminance image data and generates the determination imaged data by generating small-angle detection image data based on the small-angle maximum luminance image data, generating medium-angle detection image data based on the medium-angle maximum luminance image data, and generating large-angle detection image data based thereon the large-angle maximum luminance image data, and the defect determination part determines the presence of a defect in the inspection target area based on each of the small-angle determination image data, medium-angle determination image data, and large-angle determination image data.
[0018] According to an eighth aspect of the present invention, in the defect inspection apparatus according to any one of the first to seventh aspects, the determination image generation part generates the determination image data as binarized data, and in the case that the determination image data is generated based on the minimum luminance image data, the defect determination part determines, then, if in the determination image data there is a dark part in an area which is greater than or equal to a predetermined first dark part threshold, that there is a defect in the inspection target area; and in the case that the evaluation image data is generated on the basis of the maximum luminance image data, the defect determination part determines if there is a bright part in the determination image data in an area which is greater than or equal to a predetermined bright part threshold, that there is a defect in the examination target area.
[0019] According to a ninth aspect of the present invention, in the defect inspection apparatus according to the eighth aspect, in the case that the ceramic body is a sealed honeycomb structural body and the inspection target surface is an end face of the sealed honeycomb structural body, the evaluation image generating part generates the evaluation image data based on the minimum luminance image data and the defect determining part determines a dark one even if in the determination image data in which a dark part corresponding to an opening in the inspection target area is excluded and a light part corresponding to any connection part and a dark part corresponding to the outside of the ceramic body, if any, are additionally excluded part is present on an area which is greater than or equal to a predetermined second dark part threshold, that a defect is present in the examination target area.
[0020] According to a tenth aspect of the present invention, in the defect inspection apparatus according to any one of the third to fifth aspects, the plurality of first lighting elements and the plurality of second lighting elements are held by a single holding body, and the plurality of first lighting elements are arranged in a plane and are the plurality of second lighting elements arranged in a different plane.
[0021] According to an eleventh aspect of the present invention, in the defect inspection apparatus according to the sixth or seventh aspect, the plurality of small-angle lighting elements each include at least two individually dimmable dimming units.
[0022] According to a twelfth aspect of the present invention, in the defect inspection apparatus according to the eleventh aspect, the plurality of small-angle lighting elements, the plurality of middle-angle lighting elements and the plurality of large-angle lighting elements are held by a single holding body and are the plurality of small-angle lighting elements, the plurality of medium-angle lighting elements, and the plurality of large-angle lighting elements are arranged on respective planes different from each other.
[0023] According to a thirteenth aspect of the present invention, in the defect inspection apparatus according to any one of the first to twelfth aspects, the plurality of lighting elements included in each of the one or more lighting parts are eight lighting elements.
[0024] A method of inspecting for the presence of a defect on an outer surface of a ceramic body according to a fourteenth aspect of the present invention includes: a placing step of placing a ceramic body as an inspection target on a predetermined table; an image pickup step of generating a plurality of pieces of pickup image data by picking up, by a predetermined image pickup device, an image of an inspection target area as at least a part of an inspection target surface of the ceramic body placed on the table in a normal direction of the inspection target area; an evaluation image generating step of generating evaluation image data for determining the presence of a defect in the inspection target area based on the plurality of pieces of captured image data; and a defect determining step of determining the presence of a defect in the inspection target area based on the determination image data. In the image pickup step, the plurality of pieces of pickup image data by sequentially turning on and off four or more illumination elements provided in one or more illumination parts and configured to emit illumination light at an identical irradiation angle at respective different from and equiangularly spaced from each other around the image pickup device obliquely irradiating irradiation directions arranged around, and taking an image of the inspection target area when each of the plurality of illumination elements is turned on. In the evaluation image generating step, minimum luminance image data in which a minimum value of luminance values of the plurality of pieces of captured image data is set at an identical pixel position as a luminance value at the pixel position is generated, and then the evaluation image data is generated based on the minimum luminance image data; or maximum luminance image data in which a maximum value of luminance values of the plurality of pieces of captured image data is set at an identical pixel position as a luminance value at the pixel position is generated, and the evaluation image data is generated based on the maximum luminance image data.
[0025] According to a fifteenth aspect of the present invention, in the defect inspection method according to the fourteenth aspect, the evaluation image generating step further includes a luminance correction processing step of correcting luminance values of the plurality of pieces of captured image data such that the luminance values at a previously in each of the plurality of pieces of reference part defined in the plurality of pieces of captured image data are the same in the plurality of pieces of captured image data, and in the evaluation image generating step, the minimum luminance image data or the maximum luminance image data is generated based on the plurality of pieces of captured image data whose luminance values are corrected by the luminance correction processing step .
[0026] According to a sixteenth aspect of the present invention, in the defect inspection method according to the fourteenth or fifteenth aspect, the one or more lighting parts includes: a first lighting part including, as the plurality of lighting elements, a plurality of first lighting elements each configured to that they irradiate the examination target area with illumination light at a first irradiation angle of 30° to 60° inclusive; and a second illumination part including, as the plurality of illumination elements, a plurality of second illumination elements each configured to irradiate the inspection target area with illumination light at a second irradiation angle of 60° to 85° inclusive.
[0027] According to a seventeenth aspect of the present invention, in the defect inspection method according to the sixteenth aspect, in the imaging step, a plurality of pieces of first captured image data are generated by performing imaging of the inspection target area for each of the plurality of first lighting elements when each of the plurality of first lighting elements is turned on, and a plurality of pieces of second captured image data are generated by performing imaging of the inspection target area for each of the plurality of second lighting elements when each of the plurality of second lighting elements is turned on, and the plurality of pieces of first captured image data and the plurality of pieces of second captured image data the plurality of pieces of captured image data different from each other. In the evaluation image generating step, in the case of generating the evaluation image data based on the minimum luminance image data, first minimum luminance image data in which a minimum value of luminance values of the plurality of pieces of first captured image data at an identical pixel position is set as a luminance value at the pixel position, and second minimum luminance image data in which a minimum value of luminance values of the plurality of pieces of second captured image data at an identical pixel position is set as a luminance value at the pixel position is generated as the minimum luminance image data and the evaluation image data is generated by generating first evaluation image data based on the first minimum luminance image data and generating second determination image data based on the second minimum luminance image data; and in the case of generating the determination image data on the basis of the maximum luminance image data, first maximum luminance image data in which a maximum value of luminance values of the plurality of pieces of first captured image data at an identical pixel position is set as a luminance value at the pixel position, and second maximum luminance Image data in which a maximum value of luminance values of the plurality of pieces of second captured image data is set at an identical pixel position as a luminance value at the pixel position is generated as the maximum luminance image data and the evaluation image data is generated by generating first evaluation image data on the basis of the first maximum luminance image data and generating second determination image data based on the second maximum luminance image data. In the defect determination step, the presence of a defect in the inspection target area is determined based on both the first determination image data and the second determination image data.
[0028] According to an eighteenth aspect of the present invention, in the defect inspection method according to the sixteenth aspect, the number of the plurality of first lighting elements is equal to the number of the plurality of second lighting elements, the first and second lighting elements are arranged so that a vertical line containing an irradiation direction of each of the first lighting elements plane always contains an irradiation direction of one of the second illumination elements, and then the plurality of first illumination elements emit illumination light belonging to a first wavelength band and the plurality of second illumination elements emit illumination light belonging to a second wavelength band different from the first wavelength band, each pair of the first and the second lighting elements having irradiation directions contained in an identical vertical plane are turned on and off simultaneously in the image pickup step generates the plurality of pieces of captured image data by capturing an image of the inspection target area when each pair of the first and second lighting elements is turned on, the defect inspection method further includes a separation image generation step of generating a plurality of pieces of first separation image data and a plurality of pieces of second separation image data by performing color separation of each one of the plurality of pieces of captured image data based on the first wavelength band and the second wavelength band in the evaluation image generating step, in the case of generating the evaluation image data based on the minimum luminance image data, first minimum luminance image data, in which a minimum value of luminance values of the plurality of pieces of first separation image data at an identical pixel position is set as a luminance value at the pixel position, and second e minimum luminance image data in which a minimum value of luminance values of the plurality of pieces of second separation image data at an identical pixel position is set as a luminance value at the pixel position is generated as the minimum luminance image data and the evaluation image data is generated by generating first evaluation image data based on the first minimum luminance - generates image data and generates second determination image data on the basis of the second minimum luminance image data; and in the case of generating the determination image data on the basis of the maximum luminance image data, first maximum luminance image data in which a maximum value of luminance values of the plurality of pieces of first separation image data at an identical pixel position is set as a luminance value at the pixel position, and second maximum luminance Image data in which a maximum value of luminance values of the plurality of pieces of second separation image data is set at an identical pixel position as a luminance value at the pixel position is generated as the maximum luminance image data and the evaluation image data is generated by generating first evaluation image data on the basis of the first maximum luminance image data and generating second determination image data based on the second maximum luminance image data, and in the defect determination step, the presence of a defect in the inspection target area is determined on the basis of age of both the first determination image data and the second determination image data is determined.
[0029] According to a nineteenth aspect of the present invention, in the defect inspection method according to the fourteenth or fifteenth aspect, the one or more lighting parts includes: a small-angle lighting part including, as the plurality of lighting elements, a plurality of small-angle lighting elements, each configured to irradiate the inspection target area with illumination light at an irradiation angle of 5° to 30° inclusive; a middle-angle illumination part including, as the plurality of illumination elements, a plurality of middle-angle illumination elements each configured to irradiate the inspection target area with illumination light at an irradiation angle of 30° to 60° inclusive; and a large-angle illumination part including, as the plurality of illumination elements, a plurality of large-angle illumination elements each configured to irradiate the inspection target area with illumination light at an irradiation angle of 60° to 85° inclusive.
[0030] According to a twentieth aspect of the present invention, in the defect inspection method according to the nineteenth aspect, in the imaging step, a plurality of pieces of small-angle captured image data are generated by performing image capture of the inspection target area for each of the plurality of small-angle illumination elements when each of the plurality of small-angle lighting elements is turned on, a plurality of pieces of medium-angle pickup image data are generated by performing imaging of the examination target area for all of the plurality of medium-angle lighting elements when each is off the plurality of medium-angle lighting elements is turned on, and a plurality of pieces of large-angle pickup image data are generated by performing imaging of the examination target area for all of the plurality of large-angle lighting elements when each is off the plurality of large-angle lighting elements is turned on, the plurality of pieces of small-angle captured image data, the plurality of pieces of medium-angle captured image data, the plurality of pieces of large-angle captured image data, the plurality of pieces are different from each other Captured image data, in the evaluation image generation step, in the case of generating the evaluation image data based on the minimum luminance image data, small angle minimum luminance image data in which a minimum value of luminance values of the plurality of pieces of small angle captured image data at an identical pixel position is set as a luminance value at the pixel position, middle-angle minimum luminance image data in which a minimum value of luminance values of the plurality of pieces of middle-angle captured image data is set at an identical pixel position as a luminance value at the pixel position, and Large-angle minimum luminance image data in which a minimum value of luminance values of the plurality of pieces of large-angle captured image data is set at an identical pixel position as a luminance value at the pixel position is generated as the minimum luminance image data and becomes the determination image data by generating small-angle detection image data based on the small-angle minimum luminance image data, generating medium-angle detection image data based on the medium-angle minimum luminance image data, and generating large-angle detection image data based on the large-angle - generated minimum luminance image data; and in the case of generating the determination image data based on the maximum luminance image data, small-angle maximum luminance image data in which a maximum value of luminance values of the plurality of pieces of small-angle captured image data at an identical pixel position as a luminance value at the pixel position is set, medium-angle maximum luminance image data in which a maximum value of luminance values of the plurality of pieces of medium-angle captured image data is set at an identical pixel position as a luminance value at the pixel position, and large-angle maximum luminance image data, in which a maximum value of luminance values of the plurality of pieces of large-angle captured image data at an identical pixel position is set as a luminance value at the pixel position is generated as the maximum luminance image data and the determination image data is generated by generating small-angle determination ng image data based on the small-angle maximum luminance image data, generating medium-angle detection image data based on the medium-angle maximum luminance image data, and generating large-angle detection image data based on the large-angle maximum luminance image data is generated and in the defect determining step, the presence of a defect in the inspection target area is determined based on each of the small-angle detection image data, medium-angle detection image data and large-angle detection image data.
[0031] According to a twenty-first aspect of the present invention, in the defect inspection method according to any one of the fourteenth to twentieth aspects, the evaluation image data is generated as binarized data in the evaluation image generation step, and is generated in the defect determination step in the case that the evaluation image data is generated based on the minimum luminance image data determines that there is a defect in the inspection target area if, in the determination image data, there is a dark part in an area that is greater than or equal to a predetermined first dark part threshold; and in the case that the evaluation image data is generated based on the maximum luminance image data, it is determined that there is a defect in the inspection target area when, in the evaluation image data, there is a bright part in an area which is greater than or equal to a predetermined brightness -Part threshold is.
[0032] According to a twenty-second aspect of the present invention, in the defect inspection method according to the twenty-first aspect, when the ceramic body is a sealed honeycomb structural body and the inspection target surface is an end face of the sealed honeycomb structural body, in the evaluation image generating step, the evaluation image data is generated based on the minimum luminance image data and is generated in the defect determination step, even if in the determination image data in which a dark part corresponding to an opening in the inspection target area is excluded and a light part corresponding to any connecting part and a dark part corresponding to the outside of the ceramic body, if any, are additionally excluded, a dark one portion is present in an area greater than or equal to a predetermined second portion-dark threshold determines that a defect is in the inspection target area present.
[0033] According to a twenty-third aspect of the present invention, in the defect inspection method according to the nineteenth or the twentieth aspect, the plurality of small-angle lighting elements each includes at least two individually dimmable dimming units and becomes a luminance difference corresponding to a difference between distances of the plurality of small-angle lighting elements in a capturing range of the image capturing device by individually dimming the at least two dimming units in advance before image capturing by the predetermined image capturing device in the image capturing step.
[0034] According to a twenty-fourth aspect of the present invention, in the defect inspection method according to any one of the fourteenth to twenty-third aspects, the plurality of lighting elements provided in each of the one or more lighting parts consists of eight lighting elements.
[0035] According to the first to twenty-fourth aspects of the present invention, a defect present on an outer surface of a ceramic body, which needs to be detected, can be reliably detected without misrecognizing an irregularity of a normal ceramic surface as a defect.
[0036] In particular, according to the fifth to eighteenth aspects, the imaging time can be shortened compared to a case in which the imaging is performed while the plurality of first lighting elements and the plurality of second lighting elements are sequentially turned on.
[0037] In particular, according to the sixth, seventh, eleventh, twelve, nineteen, twenty, and twenty-three aspects, the inspection of even a large inspection target area can be performed accurately against a case of two lighting parts. character list figure 1 is an external perspective view of a honeycomb structural body 1 . figure 2 is a partially enlarged schematic drawing of an end face 1a of the honeycomb structural body 1 . figure 3 is a perspective view showing a defect which may be on the end face 1a occurs, and a normal surface irregularity ns occurring on a normal ceramic surface 6is present and does not cause a problem in terms of product specifications, illustrated schematically. figure 4 is a plan view showing a situation where a defect on the ceramic surface 6 is formed, illustrated schematically. figure 5 is a drawing showing a situation in which the end face 1a of the honeycomb structural body 1 is irradiated with illumination light in several directions, illustrated schematically. figure 6 is a drawing for describing the influence of a difference in irradiation angle of illumination light on defect detection. figure 7 is a block diagram showing the configuration of a defect inspection device 1000 according to a first configuration aspect. figure 8 is a bottom view of a main part of an image pickup execution part 100 . figure 9 is a sectional view along the line A1 - A1' in figure 8th. figure 10 is a flowchart showing the flow of the inspection by the defect inspection apparatus 1000 image pickup processing performed for defect inspection. figure 11 is a flowchart showing the flow of the first determination processing. figure 12 is a drawing for describing the first half of the first determination processing. figure 13 is a drawing for describing the second half of the first determination processing. figure 14 is a flowchart showing the flow of second determination processing. figure 15 shows a list of detection examples by the first detection processing and the second detection processing for the normal surface irregularity ns of the ceramic surface 6 , a crack df1, and a hole df3, and comprehensive investigation contents based on the investigation examples. figure 16 is a view exemplifying various types of through the process of defect inspection on the end face 1a of the honeycomb structural body 1 generated images. figure 17 is a view exemplifying various types of through the process of defect inspection on the end face 1a of the honeycomb structural body 1 generated images. figure 18 is a view exemplifying various types of through the process of defect inspection on the end face 1a of the honeycomb structural body 1 generated images. figure 19 is a block diagram showing the configuration of a defect inspection apparatus 2000 according to a second configuration aspect. figure 20 is a flowchart showing the flow of the defect inspection by the defect inspection apparatus 2000 performed image pickup processing. figure 21 is a flowchart showing the flow of color separation and reconstruction processing. figure 22 is a drawing to show a modification of the arrangement and configurations of a first lighting part 120 and a second lighting part 130 . figure 23 is a drawing to show a modification of the arrangement and configurations of the first lighting part 120 and the second lighting part 130 . figure 24 is a diagram showing a near an outer wall 2 of the honeycomb structural body 1 when irradiating the front 1a of the honeycomb structural body 1 Fig. 12 shows image IM7 taken with illumination light Ls in an oblique direction at a predetermined angle. figure 25 is a drawing for describing the influence of the difference in the distance from an illuminating element to an irradiated position. figure26 is a drawing for describing the influence of the relationship between the irradiation angle of illumination light and the angle of view on a captured image. figure 27 is a block diagram showing the configuration of a defect inspection apparatus 3000 according to a second embodiment. figure 28 is a bottom view of a main part of the image pickup execution part 100 . figure 29 is a sectional view along the line A4 - A4' in figure 28 figure 30 is a drawing for describing the relationship between one of a small-angle lighting part 115 emitted illuminating light and the viewing angle. figure 31 is a diagram for describing the effect of dimming each dimming unit. figure 32 is a flowchart showing a schematic flow of the defect inspection with the defect inspection apparatus 3000 performed image pickup processing. figure 33 is a flowchart showing a specific flow of image pickup processing. figure 34 is a flowchart showing a schematic flow of a defect inspection using the defect inspection apparatus 3000 by means of luminance correction processing 334 performed luminance correction processing. figure 35 is a diagram exemplifying the processing content of the luminance correction processing. figure 36 is a flowchart showing the flow of processing performed by a defect determination part 240 investigation processing performed. Description of Embodiments First embodiment Honeycomb structural body
[0038] First, a honeycomb structural body having an end face as a defect inspection target in the present embodiment will be described below. figure 1 is an external perspective view of a honeycomb structural body 1 . figure 2 is a partially enlarged schematic drawing of an end face 1a of the honeycomb structural body 1 .
[0039] The honeycomb body 1 is a ceramic structural body (ceramic body) with a cylindrical shape, which contains inside what is called a honeycomb structure. The honeycomb body 1 contains a large number of quadrangular-prismatic (square in cross-section) honeycombs 3 , which of an outer wall forming the cylindrical shape 2 are surrounded. Every honeycomb 3 is through a partition 4 divided (see figure 2(a)) and extends in the direction (axial direction) of the central axis of the honeycomb structural body 1 . However, the honeycomb can 3 also have a slanted prism shape, the longitudinal direction thereof with respect to the central axis of the honeycomb structural body 1 is skewed. In any case, the honeycomb 3 in a two-dimensional square grid pattern on the face 1a of the honeycomb structural body 1 arranged. In the present specification, are sections of the honeycomb structural body 1 and the honeycomb 3 to the central axis of the honeycomb structural body 1 orthogonal sections unless otherwise noted.
[0040] For example, the outer wall 2 the partition wall has a thickness of approximately 100 μm to 1500 μm 4 a thickness of about 150 µm to 400 µm and is the distance of the partition wall 4 , showing the size of the honeycomb 3 defined, approximately 1.0 mm to 2.5 mm. The length in the axial direction is about 100 mm to 300 mm, and the radius of a section orthogonal to the axial direction (cross-sectional radius) is about 100 mm to 200 mm.
[0041] More specifically, contain the honeycomb 3 a first honeycomb 3a with an opening at the front 1a and a second honeycomb 3b , which at the front 1a with a clasp 5 provided (one through the clasp 5 has a blocked opening). The first honeycomb 3aand the second honeycomb 3b are arranged alternately (in a checkerboard pattern). On the other front 1b is the first honeycomb 3a closed and is the second honeycomb 3b open. In the following description, the opening of the first honeycomb 3a at the front 1a also simply as the first honeycomb 3a designated.
[0042] The honeycomb body 1 is a fired body of ceramic (e.g. cordierite, SiC or aluminum oxide). Usually, the honeycomb structural body is obtained 1 by molding a ceramic powder as a base material of the honeycomb structural body by kneading 1 for example, clayey raw earth obtained with an organic binder and water by extrusion molding, firing a honeycomb shaped body (ceramic shaped body) thus obtained to preliminarily produce a honeycomb fired body without a shutter, and then subjecting the honeycomb fired body to a sealing treatment to remove the shutter 5 on the target honeycomb 3 to build. The closure 5 is, for example, by masking an end portion of the not with a closure 5 honeycomb to be provided 3 (on the first comb 3a directed), then filling the opened honeycomb with the filler formed by immersing an end portion of the honeycomb fired body in a slurry-like filler material containing the same ceramic powder as that used for forming the honeycomb fired body, and then firing the honeycomb fired body again.
[0043] In the figure 1 and figure 2(a) is a ceramic part of the face for better understanding 1a hatched and is the closed second honeycomb 3b (especially the second honeycomb 3b partition wall 4 ) shown with a dashed line. However, the locked second honeycomb 3b in practice from the outside on the (defect-free) front side 1a not visually recognizable. In the actual face 1a as in figure 2(b) is visually recognizable that the first honeycomb 3a in a square grid pattern on a ceramic surface hatched in the figure 6 are arranged.
[0044] the figure 3 and figure 4 are views for describing a defect possibly found on the face 1a of the honeycomb structural body formed as described above 1 occurs. A crack df1, a chip df2 and a hole df3, each of which is an indentation on the end face 1a represent are exemplary as defects which may be on the face 1a of the honeycomb structural body 1 occur, shown. figure 3 is a perspective view showing these defects and normal surface irregularities ns present on the (defect-free) normal ceramic surface 6 are present and do not cause a problem in terms of product specifications, illustrated schematically, and figure 4 is a plan view showing them on the ceramic surface 6 formed defects exemplified.
[0045] the inside figure Crack df1 shown in FIG. 3(a) is one on the ceramic surface 6 for example, crack (gap) formed due to shrinkage of the honeycomb fired body during firing. The formed crack df1 has a width of about 100 µm to 300 µm and a depth of about 160 µm to 1000 µm. The crack df1 is probably from the opening (in other words, an end part of the partition wall 4 ) of the first comb 3a on the ceramic surface 6 starting formed as in figure 4 and is sometimes from a first comb 3a up to another first honeycomb 3a educated.
[0046] In the figure 3(b) spalling shown df2 is, for example, an indentation formed when part of the ceramic surface is fired or after fired 6 missing (drops off). The formed spall df2 has a width of about 380 µm to 500 µm and a depth of about 200 µm to 1000 µm.
[0047] This in figure 3(c) hole df3is a cavity formed due to a factor such as local abnormal deformation of the ceramic surface 6 formed during burning. The formed hole df3 has a width of about 700 µm to 1500 µm and a depth of about 350 µm to 2000 µm.
[0048] figure 4 exemplifies a case in which the chipping df2 to the first honeycomb 3a at the front 1a is subsequently formed and the hole df3 in one of the first comb 3a separate part (part in which the closure 5 provided) of the ceramic surface 6 is formed, but the actual formation aspect of the spalling df2 and the hole df3 is not limited to this. For example, the hole df3 in some cases to the first comb 3a subsequently formed.
[0049] Generally speaking, the crack points df1 versus flaking df2 and the hole df3 a large ratio of depth to width, although the fissure df1 , the chipping df2 and the hole df3 are concave parts. The chipping df2 and the hole df3 have different formation factors, but are approximately the same size and need not be distinguished from each other in the defect examination to be described later. It is very important that the (defect-free) normal ceramic surface 6 , which shows the surface irregularities ns as in figure 3(d) with a depth of about 40 µm to 300 µm at intervals between the convex parts of about 50 µm to 500 µm, not erroneously as the spalling df2 and the hole df3 is recognized because such normal surface irregularities ns cause no problem in terms of product specifications.
[0050] Details of the examination for such a defect, which is on the front side 1a occurs are described below. Basic concept of defect investigation
[0051] First, the basic concept of the defect inspection performed in the present embodiment will be described below. The defect inspection performed in the present embodiment is aimed at the end face 1a of the honeycomb structural body 1 having the configuration described above, and is a schematic inspection for the presence of a defect utilizing the fact that when the end face is irradiated 1a with illumination light in an oblique direction while a defect on the face 1a exists, a shadow area (area in which the luminance is low compared to the periphery) is formed at the occurrence position, and has characteristics regarding the pattern of irradiation with illuminating light and the pattern of forming an image for determination.
[0052] figure 5 is a drawing showing the situations when irradiating the front side 1a of the honeycomb structural body 1 schematically illustrated with illumination light in several directions.
[0053] figure Fig. 5(a) is a schematic plan view of end face irradiation 1a with illumination light La in an oblique direction in a state where the honeycomb structural body 1 is arranged so that the end face 1a is essentially horizontal, and figure 5(b) is a schematic sectional view of a section including the irradiation direction of the illumination light La. In this case, if a defect (concave part) df4 as in figure 5(b) shown at the front 1a present, from most of the face 1a and the defect df4 an irradiation area RE1a irradiated with the illumination light La, but becomes, depending on the shape (width and depth) of the defect df4 and the irradiation angle (angle of irradiation direction with respect to a horizontal plane) of the illumination light La from the vicinity of a slanting surface on the left side in the defect df4 a shadow area RE2a , in which the illumination light la not noticeable.
[0054] Corresponding isfigure 5(c) is a schematic plan view when irradiating the end face 1a of the honeycomb structural body 1 with illumination light Lb and is figure 5(d) shows the irradiation direction of the illumination light Lb containing schematic sectional view, in which the honeycomb structural body 1 just like in the figure 5(a) or figure 5(b). figure Fig. 5(e) is a schematic plan view of end face irradiation 1a with illumination light Lc , and figure 5(f) is a schematic sectional view including the irradiation direction of the illumination light Lc. figure Fig. 5(g) is a schematic plan view of end face irradiation 1a with illumination light Ld, and figure 5(h) is a schematic sectional view including the irradiation direction of the illumination light Ld. However, the illumination lights have la , Lb , Lc and Ld same irradiation angles, irradiation directions lying at angular distances of 90° from each other in a horizontal plane and identical irradiation areas.
[0055] When irradiated with the illuminating light Lb arises similarly to the case of irradiation with the illumination light la , while from most of the end face 1a and the defect df4 one with the illumination light Lb irradiated irradiation area RE1b will, in part of the defect df4 on the back in the drawing plane a shadow area RE2b , in which the illumination light Lb not noticeable, which is not expressly shown in the figure.
[0056] When irradiated with the illumination light, Lc becomes while from most of the face 1a and the defect df4 one with the illumination light Lc irradiated irradiation area RE1c becomes, from the vicinity of an inclined surface of the defect df4 on the right side in the drawing plane a shadow area RE2c , in which the illumination light Lc not noticeable.
[0057] When irradiated with the illuminating light Ld arises while from most of the end face 1a and the defect df4 one with the illumination light Ld irradiated irradiation area RE1d becomes, in the vicinity of an inclined surface of the defect df4 on the front side in the drawing plane a shadow area in which the illuminating light Ld not noticeable, which is not expressly shown in the figure.
[0058] As just described, when irradiating the front side 1a , at which the defect df4 is present, with illumination light in oblique directions different from each other, the positions and shapes of respective ones for the defect df4 shadow areas formed are different from each other and in no case correspond to the entire defect df4 .
[0059] In other words, however, the fact that the positions and shapes of respective shadow areas are different suggests that the shadow areas contain information about mutually different parts of the defect df4 deliver. From this point of view, in the cases of figure 5(b), figure 5(d), figure 5(f) and figure 5(h) shadow areas formed practically superimposed on each other to figure 5(i) to obtain. In this case, the irradiation area RE1 just a part other than the defect df4 and is the defect df4 a shadow area RE2 In its entirety. In other words, the shadow area RE2 is formed in a size close to the actual size of the defect.
[0060] This means that when subsequently take a picture of the end face 1a is taken while obliquely irradiating it with illumination light in mutually different directions as in FIGS figure 5(a), figure 5(c), figure 5(e) and figure5(g), a synthesized image is generated by synthesizing the captured images so that shadow areas are superimposed on each other, and the presence of a defect is determined based on the synthesized image, the determination certainty against a case where the determination performed by simply using images obtained upon irradiation with illumination light in an oblique direction increases.
[0061] figure 5 exemplifies an aspect in which illumination light is irradiated in four directions at 90° to each other in a horizontal plane, but this is only an example and illumination light irradiation can be performed in a larger number of directions.
[0062] To be sure, in an aspect in which a plurality of illumination light beams are emitted in directions different from each other at the same time, such as an aspect in which, for example, the illumination light La and the illumination light Lc, which are emitted from positions opposite to each other, are emitted simultaneously , a place which is otherwise a shadow area when irradiated with one of the illumination light beams is irradiated with the other illumination light so that no shadow area is formed, and thus the aspect achieves no effect of increasing the reliability of flaw detection based on a shadow area . In other words, in the present embodiment, it is technically reasonable to emit illumination light one by one in a plurality of directions different from each other to obtain each image.
[0063] figure 6 is a drawing for describing the influence of a difference in irradiation angle of illumination light on defect detection. Typically, when an area where an irregularity is present is obliquely irradiated with illumination light, a shadow area is less likely to be formed since the irradiation angle is larger and the depth of the irregularity is shallower.
[0064] For example, even if the irradiation of the surface irregularities ns than one on the ceramic surface 6 present normal irregular part with illumination light LI with a relatively small angle of irradiation allows part a of the same to function as a shadow area as in figure 6(a), no shadow region is formed in some cases when the same surface irregularities ns as those in FIG figure 6(a) with illumination light Lh with an irradiation angle larger than that of the illumination light LI is to be irradiated as in figure 6(b).
[0065] the figure 6(c) and figure 6(d) show situations where a part with a crack df5 a width equal to the distance between the convex parts of the ceramic surface 6 present surface irregularities ns and a depth larger than the irregularity depth of the surface irregularities ns with the illumination light LI or the illumination light Lh is irradiated.
[0066] In such a case, as in figure 6(c), a part b as a part of the crack df5 upon irradiation with the illuminating light LI a shadow area and is also in some cases, as in figure 6(d), a part c as part of the crack df5 when irradiated with the illumination light Lh, a shadow area although narrower than the area b is.
[0067] When the defect determination based on an image of the end face obtained by irradiation with the illumination light LI 1a is performed, it may be incorrectly determined that there is a defect at the position of a shadow region formed at the normal surface irregularities ns. Consequently, irradiation with illumination light having a relatively large irradiation angle is like the illumination light Lh preferable to reliable only the crack df5and to avoid false detection of the surface irregularities ns as a defect.
[0068] However, a chip and a hole as defects each having a smaller depth and a larger width than a crack tend not to be recognized at a large irradiation angle. Thus, in the present embodiment, reliable determination is achieved by selectively using the irradiation angle of the illumination light and the presence of a defect based on a predetermined threshold according to characteristics of a face in an image 1a appearing dark part is determined for each angle. Defect inspection device and processing for defect inspection
[0069] A defect inspection device configured to actually perform a defect inspection based on the above-described concept and a specific flow of the defect inspection processing performed by the defect inspection device will be described below. The defect inspection device according to the present embodiment is roughly divided into two configuration aspects, and the contents of the processing for defect inspection differ depending on the configuration aspect of the defect inspection device. Each aspect is described below in turn. First configuration aspect defect inspection device
[0070] figure 7 is a block diagram showing the configuration of a defect inspection apparatus 1000 according to a first configuration aspect of the present embodiment. The defect inspection device 1000 mainly includes a table T on which the honeycomb structural body 1 to set as an examination target, an image pickup execution part 100 , which is configured to do so, while the honeycomb structural body placed on the table T is irradiated 1 to carry out the image recording with illumination light, and a controller 200 , which is configured to control the image pickup executive 100 and the defect determination based on a by the image pickup execution part 100 obtained captured image.
[0071] The image acquisition execution part 100 mainly includes a camera (for example, a CCD camera) 110 for taking an image of the honeycomb structural body placed on the table T 1 to record, an image recording control part 111 as a control part (camera control) for image recording by the camera 110 to control, a first lighting part 120 and a second lighting part 130 , which are each configured, the honeycomb structural body 1 to irradiate with illumination light, and a moving mechanism 140 , which is configured for the image pickup execution part 100 relative to that on the table T laid honeycomb structure body 1 to move.
[0072] figure 8 is a bottom view of a main part of the image pickup execution part 100 (Drawing in which the image pickup execution part 100 is viewed from bottom to top in the vertical direction), and figure 9 is a sectional view along the line A1 - A1' in figure 8. The cut A1 - A1' in figure 8 is an optical axis CX the camera 110 containing vertical section and is a plane of symmetry of a first lighting element 121a , a first lighting element 121e , a second lighting element 131a and a second lighting element 131e , which are to be described later.
[0073] However shows figure 9 for a better understanding also on the table T, which in figure 9 is not shown laid honeycomb structural body 1 . the figure 8 and figure 9 contain right-hand xyz coordinates with a z-axis direction along the vertical direction. In figure8, an x-axis direction is along the right-left direction in the drawing plane, and a y-axis direction is along the up-down direction in the drawing plane. Accordingly is figure 9 showing the cut A1 - A1' in figure 8 shows a zx sectional view.
[0074] Upon inspection, the honeycomb structural body 1 like that on the table T (not shown) laid that end face 1a as the inspection target surface is a horizontal top as in figure 9 shown. In the image pickup execution part 100 is to obtain a shooting target in the vertical downward direction, the camera 110 placed in a position where its objective lens faces vertically downward, and is its optical axis CX aligned in the vertical direction. Accordingly, the camera can 110 an image acquisition in a at an intersection point P between the optical axis CX and the front side 1a centered predetermined area.
[0075] Such a configuration in which the end face 1a is a horizontal top and the optical axis CX of the camera 110 aligned in the vertical direction means that the camera 110 a picture of the front 1a as an inspection target surface in the normal direction of the face 1a picks up.
[0076] The image acquisition control part 111 is on camera 110 attached and gives the camera 110 a recording instruction and guides you through the image recording by means of the camera 110 generated recording image data to the controller 200 further.
[0077] As in figure 8 are shown in the image pickup execution part 100 the first lighting part 120 and the second lighting part 130 provided so that the camera arranged in such a manner 110 surround. The first lighting part 120 and the second lighting part 130 are both at a bottom of the image pickup execution part 100 forming holding body 101 arranged, and at least when taking pictures is the camera 110 into one in the holding body 101 intended opening 102 deployed.
[0078] More specifically, the first part includes lighting 120 eight first lighting elements 121 ( 121a , 121b , 121c , 121d , 121e , 121f , 121g and 121h ) as in figure 8 shown. These eight first lighting elements 121 each have a longitudinal direction along the horizontal direction and are arranged so as to have angular intervals of 45° from each other in a horizontal plane, and are each in an inclined position on the holding body 101 fastened like the one in figure 9 exemplified first lighting elements 121a and 121e . More specifically, each first lighting element 121 an irradiation direction D1 at a predetermined angle θ1 with respect to the horizontal plane and falls the intersection between an optical axis L1 same and the front side 1a with the point of intersection P between the optical axis CX the camera 110 and the front side 1a together. Accordingly, the eight first lighting elements 121 substantially identical areas centered on the intersection P with illumination light at the same irradiation angle θ1 in mutually different directions at angular intervals of 45° from each other in the horizontal plane.
[0079] figure 9 shows an example of a situation in which the irradiation direction D1 ( D1a or D1e ) each of the two oppositely arranged first lighting elements 121a and 121e from the eight first lighting elements 121 has the predetermined angle θ1 with respect to the horizontal plane and the intersection point between the optical axis L1 ( L1a or L1e ) and the front side 1awith the point of intersection P between the cameras 110 and the front side 1a coincides.
[0080] The second lighting part 130 contains eight second lighting elements 131 ( 131a , 131b , 131c , 131d , 131e , 131f , 131g and 131h ) as in figure 8 shown. These eight second lighting elements 131 each have a longitudinal direction along the horizontal direction and are arranged so as to have angular intervals of 45° from each other in a horizontal plane, and are each in an inclined position on the holding body 101 fastened like the one in figure 9 exemplified second lighting elements 131a and 131e . More specifically, every second lighting element has 131 an irradiation direction D2 and a predetermined angle θ2, which is larger than the angle θ1, with respect to the horizontal plane and also falls the intersection point between an optical axis L2 and the front side 1a with the intersection P between the camera 110 and the front side 1a together. Accordingly, the eight second lighting elements 131 irradiate substantially identical areas centered on the intersection P with illumination light at the same irradiation angle θ2 in directions different from each other at angular intervals of 45° from each other in the horizontal plane.
[0081] figure 9 exemplifies a situation in which the irradiation direction D2 ( D2a or D2e ) each of the two oppositely arranged second lighting elements 131a and 131e from the eight second lighting elements 131 has the predetermined angle θ2 with respect to the horizontal plane and the intersection point between the optical axis L2 ( L2a or L2e ) and the front side 1a with the point of intersection P between the cameras 110 and the front side 1a coincides.
[0082] The irradiation angle θ1 of each first lighting element 121 is preferably 30° to 60° and is, for example, 45°. The irradiation angle θ2 of each second lighting element 131 is preferably 60° to 85° and is, for example, 75°. When the irradiation angle θ1 of the first lighting element 121 is less than 30°, the size of the normal surface irregularity ns of the ceramic surface increases 6 occurring shadow region and the surface irregularity ns is highly likely to be erroneously recognized as a defect, which is not desirable. When the irradiation angle θ2 of every second lighting element 131 is larger than 85°, a shadow area is unlikely to be formed at a defect and the defect is highly unlikely to be recognized, which is undesirable. The difference between θ1 and θ2 is preferably at least 15° in order to more reliably detect various types of defects having different shapes from each other.
[0083] Preferred should be a bar lighting in which a large number of LED elements are lined up in a rectangular shape as both the first lighting element 121 as well as the second lighting element 131 be used. The wavelength of each illumination light is not particularly limited, but white light or single color light can be used and the wavelengths of the illumination light of the first illumination element 121 and the second lighting element 131 can be different. Each LED element can have a maximum beam angle FWHM of approximately 5° to 30°.
[0084] However, the number is in the first lighting part 120 intended first lighting elements 121 not limited to eight, but typically m (m ≥ 4) first lighting elements 121 of the same power, each having the irradiation angle θ1, at equal angular distances from each other around the camera 110 positions arranged around in the horizontal plane.
[0085] The number in the second lighting part is corresponding 130 provided second lighting elements 131 not limited to eight, but typically there can be n (n≥4) second lighting elements 131 of the same power, each having the irradiation angle θ2, at equal angular distances from each other around the camera 110 positions arranged around in the horizontal plane. In this aspect of configuration, m ≠ n may apply.
[0086] Such a configuration that the plurality in the first lighting part 120 intended first lighting elements 121 and the plurality in the second lighting part 130 provided second lighting elements 131 are separated from each other in the respective horizontal plane while the honeycomb structural body 1 like that on the table T lies that face side 1a as the examination target surface is in a horizontal position means that the plurality of first illumination elements 121 and the plurality of second lighting elements 131 in different to the front side 1a parallel planes are separated from each other as the inspection target surface.
[0087] The movement mechanism 140 is provided to the camera 110 and the holding body 101 , on which the first lighting part 120 and the second lighting part 130 are fixed to move. In the case that the shooting range of the camera 110 because of the resolution of the camera 110 or the like smaller than the area of the face 1a of the honeycomb structural body 1 is, the movement mechanism moves 140 the camera 110 and the holding body 101 to the next picking point each time the picture picking ends at one picking point.
[0088] The defect inspection device 1000 can also be configured so that the camera 110 and the holding body 101 are fixed and the table T moves.
[0089] The control 200 is realized by a computer such as a general-purpose personal computer. The control 200 includes an input panel composed of a mouse, a keyboard and the like 201 , through which inputting of an instruction for executing the defect inspection and setting of the conditions are performed by an operator, and a display part 202 such as a display configured to perform the menu display for the defect inspection, the inspection result display, and the like.
[0090] In addition, the controller contains 200 as an integrated control part realized by executing an operation program stored in a storage part (not shown) such as a hard disk provided in the computer by a control part (not shown) provided in the computer including a CPU, a ROM and a RAM 210 , which is configured for the operation of the entire defect inspection device 1000 to control together, a lighting control part 220 , which is configured to perform the switching operations for turning on and off (ON / OFF) the lighting in the first lighting part 120 and the second lighting part 130 to control, an evaluation image generation part 230 configured to determine the presence of a defect based on image capture by the camera 110 generated captured image data to generate determination image data, and a defect determination part 240 , which is configured to determine the presence of a defect based on the determination image data.
[0091] The integrated control part 210 synchronously controls the lighting control part 220 and that in the image pickup execution part 100 provided image pickup control part 111 in response to an execution instruction to examine from the input panel 201, an image pickup for defect inspection image data of the end face irradiated with illumination light 1a to execute.
[0092] Specifically, the lighting control part switches 220 , when a predetermined control signal from the integrated control part 210 to the lighting control part 220 is sent in the first lighting part 120 provided m first lighting elements and in the second lighting part 130 provided n second lighting elements turn on and off sequentially at a predetermined point in time for an on-time in response to the signal.
[0093] On the other hand, from the integrated control part 210 a control signal for subsequently carrying out the image recording by means of the camera 110 in synchronism with the successive turning on of the m first lighting elements 121 and the n second lighting elements 131 to the image pickup control part 111 Posted. The image acquisition control part 111 controls the camera 110 so that it performs the image pickup at a predetermined timing in response to the control signal.
[0094] The integrated control part 210 executes the instruction to move the image pickup executive 100 to the next pick-up point at the time when picture pick-up ends at a certain pick-up point. In addition, the integrated control part leads 210 processing for displaying by the defect determination part 240 generated determination result data on the display part 202 through.
[0095] The detection image generation part 230 captured by taking pictures with the camera 110 generated recording image data directly or indirectly (via the integrated control part 210 ) from the image capture control section 111 and generates detection image data. The detection image generation part 230 includes a minimum luminance imaging part 231 and a filter processing part 232 as functional components for performing the generation of the determination image data.
[0096] As described above, the first part is lighting 120 with the m (e.g. eight) first lighting elements 121 provided and will capture the image every time the first lighting elements 121 be switched on one after the other, then carried out. Thus, one obtains by an in the irradiation angle θ1 on the face 1a of the honeycomb structural body 1 Emitted illumination m pieces of captured image data (first captured image data). The second lighting part is corresponding 130 with the n (e.g. eight) second lighting elements 131 provided and will capture the image every time the second lighting elements 131 be switched on one after the other, then carried out. Thus, one obtains by an in the irradiation angle θ2 on the face 1a of the honeycomb structural body 1 emitted illumination n pieces of captured image data (second captured image data).
[0097] The m pieces of first captured image data and the n pieces of second captured image data are subjected to luminance correction as preprocessing and in the minimum luminance image generation part 231 used for minimum luminance imaging. Various types of methods can be applied to the luminance correction, and for example, luminance correction processing performed in a second embodiment to be described later can be applied to the luminance correction. The luminance correction processing can be performed using the camera 110 , the image pickup control part 111 or the minimum luminance imaging part 231 performed as long as they have a function to perform the processing. Alternatively, such as a defect inspection device 3000 according to the second embodiment to be described later, a processing part configured to perform the luminance correction processing may be provided.
[0098] The minimum luminance imaging part 231 performs the processing of generating first minimum luminance image data from the m pieces of first captured image data and generating second minimum luminance image data from the n pieces of second captured image data.
[0099] The first minimum luminance image data is image data in which B 1(x,y) is given in a following expression, in which B 1(x,y) represents the luminance value at a pixel (x,y) and B 1(x,y)i represents the luminance value at a single pixel (x, y) of the i-th first captured image data: B 1 ( x , y ) = Min { B 1 ( x , y ) 1 , B 1 ( x , y ) 2 , … B 1 ( x , y ) m }
[0100] In other words, the first minimum luminance image data is image data in which the minimum value Min{B 1(x,y)1 , B 1(x,y)2 , ..., B 1(x,y)m} of the luminance value at each pixel (x, y) in the m pieces of first captured image data is the luminance value at the pixel (x, y).
[0101] Correspondingly, the second minimum luminance image data is image data in which B 2(x,y) is given in a following expression, in which B 2(x,y) represents the luminance value at a pixel (x,y) and B 2(x,y)j represents the luminance value at a single pixel (x, y) of the j-th second sample image data: B 1 ( x , y ) = Min { B 1 ( x , y ) 1 , B 1 ( x , y ) 2 , … B 1 ( x , y ) m }
[0102] In other words, the second minimum luminance image data is image data in which the minimum value Min{B 2(x,y)1 , B 2(x,y)2 , ..., B 2(x,y)n} of the luminance value at each pixel (x, y) in the n pieces of second captured image data is the luminance value at the pixel (x, y).
[0103] When the m pieces of first captured image data and the n pieces of second captured image data include a shadow area corresponding to a defect, the first minimum luminance image data and the second minimum luminance image data represent an image in which a shadow area is practically superimposed in each piece of captured image data as an example in figure 5 shown conceptually. This means that a shadow area (area where the luminance value is low) due to a defect is enhanced in the first minimum luminance image data and the second minimum luminance image data.
[0104] On the other hand, even if the first captured image data or the second captured image data obtained by performing image capturing under a first illuminating element or second illuminating element emitting light in a certain direction, an on the ceramic surface 6 present normal surface irregularity ns contain shadow area attributable to the present normal surface irregularity ns are chips and holes in the first photographed image data or the second photographed image data obtained by performing the image pickup while irradiating the same place with light in different directions from the first illuminating element or the second illuminating element opposite to that of the Shadow region corresponding to surface irregularity ns is increased because the area of surface irregularity ns is relatively small. A priori, the shadow region corresponding to the surface irregularity ns is unlikely to be formed under the second illuminating element having a large irradiation angle.
[0105] In the defect inspection device 1000 improves the security of detecting a defect on the face 1a of the honeycomb structural body 1 increased by using the first minimum luminance image data and the second minimum luminance image data having such characteristics in determining the presence of defects.
[0106] The filter processing part 232 subjects the first minimum luminance image data and the second minimum luminance image data generated by the minimum luminance image generation part 231 were generated, various types of filter processing and generates data more suitable for determining the presence of defects. The filter processing includes the well-known image processing techniques of binarization processing, termination processing (expansion / contraction processing), and marker processing.
[0107] The filter processing part 232 generates first determination image data by subjecting the first minimum luminance image data to the filter processing, and generates second determination image data by subjecting the second minimum luminance image data to the filter processing.
[0108] Schematically, the filter processing part leads 232 performs the binarization processing on both the first minimum luminance image data and the second minimum luminance image data in which each pixel has a gradation value based on a predetermined luminance threshold, and removes a very small area of dark pixels as a noise component by forming a pixel area ( Area in which dark pixels are present throughout) from which, as a result of binarization, a dark pixel of luminance 0 undergoes finishing processing, and performs marking of the remaining dark pixel area by means of the marking processing, thereby generating the first determination image data and the second determination image data.
[0109] In addition, the filter processing part leads 232 an opening mask processing of excluding the first honeycomb 3a , which is always a shadow area in an image represented by both the first determination image data and the second determination image data, from the defect determination targets.
[0110] The defect detection part 240 determines the presence of a defect based on the first determination image data and the second determination image data. When a dark pixel region in an area which is greater than or equal to a predetermined threshold in a detection image represented by both the first detection image data and the second detection image data or in a detection image represented by subjecting the first detection image data and the second detection image data to the opening mask processing obtained determination image, the defect determination part determines 240 schematically that there is a defect at the occurrence position of the dark pixel area. image capture processing
[0111] figure 10 is a flowchart showing the flow of image pickup processing used for defect inspection in the defect inspection apparatus 1000 is performed with the configuration described above. In figure 10 and the related description, the honeycomb structural body 1 also referred to as a “workpiece” as a defect inspection target and becomes the face 1a as an inspection target surface of the honeycomb structural body 1 also referred to as a "workpiece face".
[0112] First, the work is placed in position on the table by an operator or a predetermined conveyor (placer). T laid in which its face is a top (step S1 ). After placing the workpiece, if via the input panel 201 an execution instruction for a defect inspection is given, the movement mechanism 140 driven to the image pickup execution part 100 (specifically the camera 110 and the first lighting part 120 and the second lighting part 130 supporting holding body 101 ) to move to a pick-up location (step S2 ). When the shooting range of the camera 110 smaller than the surface of the workpiece face 1a is, the inspection is performed many times, and hence the inspection target area or the shooting site in one inspection processing is a predetermined part of the face 1a .
[0113] In this case, the workpiece when it is on the table T is laid, so can be positioned or the position of the camera 110 be adjusted in the horizontal plane so that the honeycomb 3 of the workpiece (the first honeycombs 3a in appearance) in the shooting range of the camera 110 , which is defined in a rectangular shape, are lined up in the long axis and short axis directions. However, even if the row direction of the honeycomb 3with respect to the longitudinal and transverse axis directions in the shooting range of the camera 110 is a little skewed, the detection processing can be performed smoothly by making a correction in the detection processing considering the skewing as needed.
[0114] A sensor configured to sense whether the work is placed on the table T may be provided, and the integrated control part 210 sends, in response to a sensing signal from the sensor, a predetermined control signal for sequentially executing the image pickup processing and the subsequent detection processing to each component of the defect inspection apparatus 1000 send.
[0115] After the image pickup execution part 100 is positioned at the recording point, the initial value of i = 1 is set (step S3 ) and the consecutive image recording by the camera 110 performed while leading to the first lighting part 120 belonging m first lighting elements 121 be switched on one after the other.
[0116] Specifically, this becomes the first lighting part 120 associated i-th first lighting element 121 switched on (step S4 ) and takes the camera 110 captures an image of the workpiece in such a power-on state (step S5 ). The i-th first captured image data obtained by the image capture is controlled by the image capture control part 111 to the minimum luminance imaging part 231 forwarded (step S6 ) and subjected to detection processing to be described later. After completion of the image recording and the forwarding, the i-th becomes the first lighting element 121 , which is on, off (step S7 ). Alternatively, the i-th first lighting element 121 be turned off immediately after the image acquisition is complete. Alternatively, all pieces of first captured image data after completion of the image capture that was performed during the m-th first lighting element 121 was on, to the minimum luminance imaging part 231 to get redirected.
[0117] If i = m does not hold (NO in step S8 ), in other words if there is any first lighting element 121 is still to be switched on, i = i + 1 is set (step S9 ) and become step S4 and repeats the following processing.
[0118] If i = m (YES in step S8 ), in other words, if all of the m first lighting elements 121 be switched on one after the other and the image recording by means of the camera 110 is subsequently performed in each power-on state, the initial value of j = 1 is set (step S10 ) and the consecutive image recording by the camera 110 performed while going to the second lighting part 130 belonging n second lighting elements 131 be switched on one after the other.
[0119] Specifically, this becomes the second lighting part 130 associated j-th second lighting element 131 switched on (step S11 ) and takes the camera 110 captures an image of the workpiece in such a power-on state (step S12 ). The j-th second captured image data obtained by the image capture is controlled by the image capture control part 111 to the minimum luminance imaging part 231 forwarded (step S13 ) and subjected to detection processing to be described later. After completion of the image recording and the forwarding, the j-th second lighting element 131 , which is on, off (step S14 ). Alternatively, the j-th second lighting element 131 be turned off immediately after the image acquisition is complete.
[0120] If j = n does not hold (NO in step S15 ), in other words, if there is any second lighting element that is yet to be turned on, j = j + 1 is set (step S16 ) and become step S11and repeats the following processing.
[0121] If j = n (YES in step S16 ), in other words, if all of the n second lighting elements 131 be switched on one after the other and the image recording by means of the camera 110 is then performed in each power-on state, the image pickup processing ends.
[0122] When the shooting range of the camera 110 is smaller than the area of the workpiece face, becomes the image pickup execution part 100 moved to the next pickup location to step processing S3 and then repeat. discovery processing
[0123] The following is the by the defect inspection device 1000 determination processing performed to determine the presence of a defect will be described. The defect inspection device 1000 performs first determination processing based on the m pieces of first captured image data and second determination processing based on the n pieces of second captured image data. In the present configuration aspect, after the completion of the image pickup processing, the first determination processing and the second determination processing need not both be performed, but the first determination processing can be performed at a time when all pieces of first captured image data are generated by the image pickup processing, and the second determination processing can be performed at a point of time when all pieces of second captured image data are generated.
[0124] figure 11 is a flowchart showing the flow of the first determination processing. In the first determination processing, the minimum luminance image generation part first generates 231 based on the above expression ( 1 ) the first minimum luminance image data from the image pickup control part 111 forwarded m pieces of first recording image data (step S21 ).
[0125] The by the minimum luminance imaging part 231 generated first minimum luminance image data is subjected to the filter processing in the filter processing part 232 subjected (step S22 ).
[0126] Specifically, first, the well-known binarization processing is performed in which a pixel (x, y) is treated as a bright pixel of luminance 1 is set when the luminance value B 1(x,y) at which pixel (x,y) is greater than or equal to a predetermined luminance threshold, and a pixel (x,y) as a dark pixel of luminance 0 is set when the luminance value B 1(x,y) is less than the predetermined luminance threshold. The pixel set as a bright pixel at this stage is excluded from the targets of the following detection processing. Hereinafter, an area in which there are dark pixels throughout is also referred to as a dark part or a dark area.
[0127] Then, the well-known final processing (expansion / contraction processing) is performed for the dark part, so that a dark part, which is discretely present as a noise component in image data after binarization processing and has a small area area (containing a small number of pixels), is eliminated from targets of the following investigation processing is excluded.
[0128] As described above, in some cases, image data contains, as a processing target, a reference to the normal surface irregularities ns formed on the ceramic surface 6 are present and do not cause a problem in terms of product specifications, due dark part. Such a dark part has a relatively small region area and hence is mostly excluded from targets of the detection processing by the final processing.
[0129] Finally, the well-known tag processing of associating all dark parts with identification information for uniquely identifying each dark part is performed to identify each dark part remaining after the final processing.
[0130] The first determination image data obtained through the filter processing described above is subject to determination by the defect determination part 240 subjected. figure 12 is a drawing for describing the first half of the first determination processing. figure 13 is a drawing for describing the second half of the first determination processing.
[0131] figure 12(a) exemplifies a first determination image represented by the first determination image data IM1 . This in figure 12(a) first determination image shown IM1 corresponds to the in figure 4 exemplified ceramic surface 6 . In figure 12 is a dark part SD crosshatched.
[0132] In the first investigative picture IM1 , which in figure 12(a), there are several types of the dark part SD . Specifically, there are nine dark parts in total SD , including six square dark parts SD0 , dark parts SD1 and SD2 , in each of which a square dark part attached to a crack df1 or the flaking df2 as in figure 4 exemplifies the dark part to be returned, and a dark part SD3 , which is independent of a square dark part and points to the in figure 4 exemplified hole df3 is due. These nine dark parts SD are associated with identification information in advance by the tag processing.
[0133] The reason why the dark parts SD0 in the first investigation picture IM1 are present is that the first determination image data is data based on the m pieces of first captured image data included in the capturing area at the front side 1a opened first honeycomb 3a were recorded. Hence the dark parts lie SD0 also in the defect-free ceramic surface 6 respective first determination image data in a grid pattern.
[0134] After generating the first determination image data, the defect determination part leads 240 performs the determination processing as to the presence of a defect on the first determination image data.
[0135] Specifically, first the first evaluation image data is checked against a predetermined first threshold (first dark part threshold) to determine whether any of the marked dark parts included in the first evaluation image data SD occupies an area (specifically, the number of pixels contained corresponding to the area) which is greater than or equal to the first threshold (step S23 ). This determination based on the first threshold is performed to detect a defect such as the crack df1 or the flaking df2 , which is attached to the first honeycomb 3a connects as in figure 4 exemplifies to recognize.
[0136] Consequently, the first threshold is a little higher than the area (the number of pixels) of the opening of the first honeycomb 3a set. Accordingly, it is possible to place one on the first honeycomb 3a , whose opening connects to a defect, attributable dark part SD reliably and at the same time false detection of a normal dark part SD0 , which is on the normal first honeycomb 3a , the opening of which is not contiguous to a defect, as a site of defect occurrence.
[0137] For a better understanding, in figure 12(a) a first threshold region TH1 with an area corresponding to the first threshold exemplified in dashed lines as a square region. However, the first emerging area is TH1 of course not limited to a square area. figure 12(b) shows an example by extracting only dark parts SD having an area greater than or equal to the first threshold from the first determination image IM1 obtained extraction image IM2 .
[0138] Only the dark parts SD1 and SD2, each having a larger area than the first threshold region TH1 (as the dark parts, of course SD0 ) have and which by the presence of an attached to the first honeycomb 3a subsequent defect are caused in the in figure 12(b) shown extraction image IM2 before, and the dark parts SD0 are excluded. The hole df3 corresponding dark part SD3 will be in the extraction image IM2 also excluded because the area of the dark part SD3 smaller than the first threshold area TH1 is.
[0139] Even if at this stage any on the ceramic surface 6 present normal surface irregularity ns remains, such dark part is smaller than the first threshold and hence it is reliably excluded from the extraction targets, and accordingly the surface irregularity ns is not erroneously recognized as a defect.
[0140] If there is no dark part SD (like the dark parts SD1 and SD2 ) with an area greater than or equal to the first threshold (NO in step S23 ), it can at least be said that a to the first honeycomb 3a subsequent defect is not recognized (step S24 ). On the other hand, if such a dark part SD present (YES in step S23 ), determined that this form of defect on the ceramic surface 6 available (step S25 ).
[0141] The first determination image data for which no defect is detected as a result of the comparison with the first threshold becomes the one processed by the filter processing part 232 performed opening mask processing (step S26 ). In the event that a defect is detected in the first detection image data, the first detection processing may be ended as it is because it has been confirmed that the ceramic surface is present 6 there is a defect as the target of the investigation. Alternatively, the first determination image data may be a target of the following defect determination processing, and in this case, the first determination image data becomes similar to a case where no defect is detected by the filter processing part 232 performed opening mask processing.
[0142] The opening mask processing is a processing of excluding all rectangular dark parts corresponding to the first honeycomb 3a in the first detection image data correspond to from the targets of the detection processing. The arrangement position (the arrangement pitch) and the size (and the shape in a case of the first honeycomb 3a near the outer wall 2 ) the one on the front side 1a of the honeycomb structural body 1 present first honeycomb 3a are known. The filter processing part 232 performs, for example, pattern matching, honeycomb area expansion / contraction processing, and binarization processing using a maximum luminance value image, based on this information inputted in advance, to thereby detect a dark portion which appears to be the same shape as that of the first honeycomb 3a at the front 1a has been determined to be excluded from the objectives of the determination processing. This corresponds to a processing of practically attaching a mask in a rectangular dark part.
[0143] figure 13(a) shows by practically attaching a mask MS in a first determination image represented by the first determination image data IM1 lying rectangular area IM3 . In opening mask processing, the mask MS regardless of whether each part a one to the first honeycomb 3a subsequent defect corresponding part or not, practically in all in the first determination image IM1 present rectangular dark parts SD appropriate. Consequently, only a dark part remains SD1a , which is part of the dark part SD1 in the first investigation picture IM1 is and the infigure 4 exemplified crack df1 corresponds, a dark part SD2a , which is part of the dark part SD2 is and the spalling df2 corresponds, and which corresponds to the hole df3 corresponding dark part SD3 than the dark parts SD in the mask-processed image IM3 .
[0144] Then, the first evaluation image data after the opening mask processing is checked against a predetermined second threshold (second dark part threshold) to determine whether any of the dark parts remaining after the opening mask processing SD occupies an area (specifically, the number of pixels contained corresponding to the area) which is greater than or equal to the second threshold (step S27 ). This determination on the basis of the second threshold is carried out in order notably to be independent of the first honeycomb 3a present defect which was not detected by the above-described determination based on the first threshold. Consequently, the second threshold is set to a value that is sufficiently small compared to the first threshold. For example, the second threshold is preferably a little larger than the distance between the convex parts of the normal surface irregularities ns of the ceramic surface 6 set. However, since each rectangular dark part is masked, the detection target also includes a one to the first honeycomb 3a a dark part corresponding to a subsequent defect and adjacent to a rectangular dark part. Such a defect has already been detected by the determination based on the first threshold as described above, but the double detection causes no problem.
[0145] For a better understanding, in figure 13(b) a second threshold region TH2 with an area corresponding to the second threshold shown by way of example with dashed lines. More specifically, there are three second emerging regions TH2 ( TH2a , TH2b and TH2c ) are shown with different shapes from each other.
[0146] If no dark part after opening mask processing SD with an area greater than or equal to the second threshold is present in the first determination image data (NO in step S27 ), it can be said that no independent of the first honeycomb 3a existing defect is detected (step S28 ). On the other hand, if such a dark part SD present (YES in step S27 ), determines that a certain type of defect is present (step S29 ). More specifically, if no defect is detected by the comparison with the first threshold but a defect is detected by the comparison with the second threshold, it means that one of the first honeycomb 3a independent defect exists. If a defect is detected by the comparison with the first threshold and a defect is detected by the comparison with the second threshold, it is not necessarily said that a honeycomb from the first honeycomb 3a independent defect exists.
[0147] figure 14 is a flowchart showing the flow of the second determination processing performed after the first determination processing. In the second determination processing, similarly to the first determination processing, the second minimum luminance image data is determined based on the expression ( 2 ) is generated from the n pieces of second pickup image data (step S31 ), the second minimum luminance image data becomes the filter processing in the filter processing part 232 subjected (step S32 ) and the resultant second determination image data becomes the determination by the defect determination part 240 subjected. The special processing content (steps S33 until S39 ) by the defect determination part 240 is the same as the processing content (steps S23 until S29) of the first determination processing, except that a third threshold is used as the first dark-part threshold instead of the first threshold, and a fourth threshold is used as the second dark-part threshold instead of the second threshold. Thus, a detailed description regarding the second determination processing is omitted.
[0148] The second captured image data used in the second determination processing is obtained under illumination light with a large irradiation angle compared to the first captured image data used in the first determination processing. Consequently, from the first recording image data, it is unlikely that an incident occurs on the normal surface irregularity ns of the ceramic surface 6 attributable shadow area occurs in a captured image. Thus, in the second minimum luminance image data generated based on Expression (2) and the second determination image data obtained by performing filter processing on the second minimum luminance image data, it is less likely than in the first determination image data that a dark part corresponding to the surface irregularity ns exists . Thus, in the second determination processing, false detection of the normal surface irregularity ns is more reduced than that in the first determination processing.
[0149] Incidentally, by generating the second minimum luminance image data based on Expression (2), a dark part for a defect such as the crack is still likely to become df1 formed with a small width and a large depth (see figure 6). However, the necessity of re-detecting the defect already detected by the first detection processing by the second detection processing is small, and the second detection processing is very useful for detecting a defect which was not detected by the first detection processing. Therefore, the third threshold and the fourth threshold are set lower than the first threshold and the second threshold, respectively. Even if a defect like the crack df1 having a small width and a large depth is not detected by the first detection processing, such a defect can be reliably detected by performing the second detection processing.
[0150] The determination processing results of the first determination processing and the second determination processing become, as necessary, determination result data by the defect determination part 240 to the integrated control part 210 Posted. The integrated control part 210 controls the display part based on the description content of the determination result data 202 to display a defect determination result. The display can be in different types of formats. For example, only the presence of a defect in an inspection target area can be indicated, and the position of the defect can be indicated based on marker information. In addition, the size of the defect can be displayed based on the area (the number of pixels) of a dark part.
[0151] figure 15 shows a list of detection examples by the first detection processing and the second detection processing for the normal surface irregularity ns of the ceramic surface 6 , the crack df1 and the hole df3 (see for example figure 3) based on the above description and comprehensive determination contents based on the determination examples. Since the evaluation for a chip is performed in the same manner as that for a hole, Fig figure 15 representative of a case of a hole. In figure 15 "OK" means no detection as defect and "NG" means detection as defect.
[0152] More specifically shows figure 15 regarding the first detection processing, the states of the normal surface irregularity ns of the crack df1 and the hole df3 , while any of the first lighting elements 121 (the irradiation direction D1) is on, the states of normal surface irregularity ns, crack df1 and the hole df3 in a determination image represented by the first determination image data generated based on the m pieces of first captured image data, and results of determination from the states.
[0153] Correspondingly shows figure 15 regarding the second detection processing, the states of the normal surface irregularity ns, of the crack df1 and the hole df3 , while any of the second lighting elements 131 (the irradiation direction D2 ) is on, the states of normal surface irregularity ns, crack df1 and the hole df3 in a determination image represented by the second determination image data generated based on the n pieces of second captured image data, and results of determination from the states.
[0154] In addition, shows figure 15 comprehensive (final) determination results based on the results of the first determination processing and the second determination processing.
[0155] First, as for the normal surface irregularity ns, even if in a first evaluation image by synthesizing shadow parts A , which are generated when the individual first lighting elements 121 be switched on one after the other, a shadow area A' is formed, the shadow area A' normally smaller than a threshold area corresponding to the second threshold THa (obviously smaller than an area corresponding to the first threshold) and consequently it is not recognized as a defect.
[0156] Since the angle θ2 of the irradiation direction D2 of the second lighting element 131 with respect to the horizontal plane is greater than the angle θ1 the irradiation direction D1 of the first lighting element 121 with respect to the horizontal plane, a shadow area is not normally formed at the surface irregularity ns in a second evaluation image. When formed, a shadow area is smaller than a threshold area corresponding to the fourth threshold THb (Of course smaller than an area corresponding to the third threshold).
[0157] Consequently, a normal surface irregularity is not erroneously recognized as a defect by either the first determination processing or the second determination processing.
[0158] Meanwhile, what will the crack df1 as to, in the first evaluation image by synthesizing shadow parts B , which are generated when the individual first lighting elements 121 be switched on one after the other, a shadow area B' formed and there may be a case where at a relatively small size of the shadow area B' the shadow area B' smaller than the threshold area THa is. In this case, the crack df1 is not recognized as a defect by the first determination processing.
[0159] However, in a case of tear df1 shadow parts D generated when the individual second lighting elements 131 are sequentially turned on, and consequently a shadow area becomes in the second evaluation image D' educated. Consequently, when the emerging area THb is set so that the shadow area D' recognized as a defect, the shadow area D' detected as a defect by the second determination processing even if not detected as a defect by the first determination processing.
[0160] That is, the crack is determined as a defect at least by the second determination processing even if it is not determined as a defect by the first determination processing. Of course, in the event that the shadow area B' of the crack in the first detection processing is larger than the threshold area THa , the crack is recognized as a defect by the foregoing first determination processing.
[0161] Also, what will the hole df3as to, in the first evaluation image by synthesizing shadow parts C , which are generated when the individual first lighting elements 121 be switched on one after the other, a shadow area C' educated. If the threshold area THa is set so that the shadow area C' when a defect is detected, the hole becomes df3 recognized as a defect by the first determination processing.
[0162] There the hole df3 has a relatively large width, a shadow area is unlikely to be formed in the second evaluation image. Consequently, the hole df3 , although it is unlikely to be recognized as a defect by the second determination processing, is recognized as a defect by the first determination processing and is thus finally recognized as a defect with no problem. The same applies to a chipping.
[0163] Thus, according to the present aspect of configuration, a hole and a chip can be detected by the first detection processing, and a crack can be detected by the second detection processing even if not detected by the first detection processing. Moreover, a normal surface irregularity will not be mistaken for a defect. In other words, according to the present aspect of configuration, highly accurate defect detection is achieved. Sample image and effect of using minimum luminance imagery
[0164] the figure 16 to figure 18 are illustrations showing various types of lighting elements through the process of using the eight first ones 121 and defect inspection apparatus including eight second lighting elements 1000 as in figure 8 shown at the front 1a of the honeycomb structural body 1 as an examination objective carried out images generated images exemplify. All in the figure 16 to figure 18 captured images are under irradiation with the first illumination elements 121 taken at the irradiation angle θ1 of 45°.
[0165] First are in figure 16 images shown IM4a , IM4c , IM4e and IM4g captured images of first captured image data obtained while the first lighting elements 121a , 121c , 121e or. 121g near a first comb 3a be turned on. A minimum luminance image IM4 is determined by based on the four pieces of first captured image data and the four pieces of first captured image data (in figure 16 not shown), which is obtained while the remaining first lighting elements 121b , 121d , 121f or. 121h are switched on, generated first minimum luminance image data is displayed.
[0166] In figure 17 images shown IM5a , IM5c , IM5e and IM5g are recorded images of first recording image data, which can be obtained for a part of the ceramic surface 6 the front side 1a just like those of the four in figure 16 captured images shown. A minimum luminance image IM5 is determined by based on the four pieces of first captured image data and the four pieces of first captured image data (in figure 17 not shown), which is obtained while the remaining first lighting elements 121b , 121d , 121f or. 121h are switched on, generated first minimum luminance image data is displayed.
[0167] In addition, figure 18 images shown IM6a , IM6c , IM6e and IM6g recorded images of first recorded image data, which one has for the defect-free ceramic surface having only normal surface irregularities 6 just like those of the four in figure 16 captured images shown. A minimum luminance image IM6 is determined by based on the four pieces of second captured image data and the four pieces of first captured image data (in figure18 not shown) generates first minimum luminance image data, which is obtained while the remaining first lighting elements 121b , 121d , 121f or. 121h be switched on, shown.
[0168] A one to the first honeycomb 3a subsequent defect (hole) df6 corresponding dark part is in the in figure 16 shown minimum luminance image IM4 observed. The defect df6 corresponding dark part will not be in IM4g observed but is in the picture IM4c at the same position as that in the minimum luminance image IM4 observed and partially in the pictures IM4a and IM4e observed. However, the area of the dark part is in the minimum luminance image IM4 bigger than in the picture IM4c .
[0169] in the in figure 17 shown minimum luminance image IM5 becomes a one from a first honeycomb 3a to another first comb 3a extending defect (crack) df7 corresponding dark part observed. The defect df7 corresponding dark part is shown in the pictures IM5a and IM5e observed is in the pictures IM5c and IM5g but indistinct.
[0170] However, in the in figure 18 minimum luminance image shown IM6 , apart from the first combs present on four sides of the picture 3a corresponding dark parts, no clear dark part formed.
[0171] A comparison of the figure 16 and figure 17 minimum luminance images shown IM4 and IM5 with the inside figure 18 minimum luminance image shown IM6 shows that if a defect on the face side 1a exists, a dark part corresponding to the defect is clearly formed in the minimum luminance image, but no clear difference in the minimum luminance image of the ceramic surface 6 , where there is no defect and only the normal surface irregularity ns is formed, occurs.
[0172] However, for example, the pictures IM5c and IM5g in figure 16 captured images of an area where the defect df7 are present, but they do not differ significantly from the minimum luminance image IM6 in figure 18
[0173] These indicate that in the case where the presence of a defect is determined using photographed image data obtained by irradiating illumination light in a single irradiation direction, a defect may not be detected depending on the irradiation direction of the illumination light, but in the case that the presence of a defect is detected using minimum luminance image data generated from a plurality of pieces of photographed image data obtained by irradiating illumination light in irradiation directions different from each other, a defect can be reliably detected.
[0174] As described above, according to the present configuration aspect, a defect that needs to be detected can be reliably detected without misrecognizing any irregularity on a normal ceramic surface as a defect by using illumination light when detecting the presence of a defect from a plurality of pieces of ceramic minimum luminance image data generated in irradiation directions different from each other are used. Second configuration aspect
[0175] figure 19 is a block diagram showing the configuration of a defect inspection apparatus 2000 according to a second configuration aspect of the present embodiment. Components of the defect inspection device 2000 are largely the same as the components of the defect inspection apparatus 1000 according to the first configuration aspect, and hence any common parts are denoted by the same reference numerals and detailed description thereof will be omitted and mainly any differences from the defect inspection apparatus will be discussed below 1000 described.
[0176] Similar to the defect inspection device 1000 contains the defect inspection device 2000 above all: the table T , on which the honeycomb structural body 1 in a position in which the end face 1a is horizontal as the inspection target surface is to be laid; the image pickup execution part 100 , which is configured for the imaging of the face 1a of the honeycomb structural body placed on the table T 1 during frontal irradiation 1a to perform with illuminating light; and the controller 200 , which is configured to control the image pickup executive 100 and the defect determination based on a by the image pickup execution part 100 obtained captured image.
[0177] The configuration and placement of the camera 110 and the configurations and arrangement of the first lighting part 120 and the second lighting part 130 , which are in the image pickup execution part 100 are essentially the same as configurations and arrangement in the case of the figure 8 and figure 9 for a case of m=8 and n=8. That is, to obtain a shooting target in the vertical downward direction is the camera 110 placed in a posture where its lens faces vertically downward, and its optical axis CX is aligned in the vertical direction. The plurality of first lighting elements 121 and the plurality of second lighting elements are in a horizontal plane at the irradiation angles θ1 and θ2, respectively, equiangularly spaced from each other around the camera 110 arranged around.
[0178] However, in the defect inspection device 2000 n = m and is the irradiation direction D2 a second lighting element 131 inevitably in a direction of irradiation D1 a first lighting element 121 containing vertical plane.
[0179] A wavelength band (first wavelength band) to which the wavelength of the first illumination elements 121 emitted illumination light belongs, and a wavelength band (second wavelength band) to which the wavelength of the light emitted from the second illumination elements 131 emitted illuminating light are set differently from each other. This is done by using the first lighting elements, for example 121 , which are configured to emit red light, and the second lighting elements 131 , which are configured to emit blue light, or achieved by using the reverse configuration. In this case, it is defined that the red light belongs to the wavelength band of an emission wavelength of 600 nm to 800 nm and the blue light belongs to the wavelength band of an emission wavelength of 400 nm to 500 nm. Alternatively, this is also done by using the first lighting elements, for example 121 , which are configured to emit white light, and the second lighting elements 131 , which are configured to emit ultraviolet light, or by using the reverse configuration. In this case, it is defined that the white light belongs to the wavelength band of an emission wavelength of 300 nm to 800 nm and the ultraviolet light belongs to the wavelength band of an emission wavelength of 100 nm to 400 nm.
[0180] Accordingly, the camera 110 with an excellent sensitivity for the first wavelength band to which the wavelength of the first illumination elements 121 emitted illumination light belongs, and the second wavelength band to which the wavelength of the from the second illumination elements 131 emitted illuminating light is used.
[0181] In the defect inspection device 2000 according to the present aspect of configuration, a pair emits from each first lighting element 121 and corresponding second lighting element 131, their irradiation directions D1 and D2 contained in an identical vertical plane simultaneously emit illumination light in an identical direction at the different irradiation angles θ1 and θ2 around the intersection point P between the optical axis CX the camera 110 and the front side 1a around from (see figure 9). This means that in the defect inspection device 2000 according to the present configuration aspect, illumination light beams having wavelengths belonging to different wavelength bands from each other are emitted in an identical direction in a superimposed manner. In other words, the defect inspection device 2000 according to the present configuration aspect, the includes m (e.g., eight) pairs of first lighting elements 121 and second lighting elements 131 . Accordingly, the camera leads 110 performs the imaging in a state where such two illumination light beams are irradiated in a superimposed manner.
[0182] The control 200 differs from the controller 200 the defect inspection device 1000 that additionally a separation image generation part 233 in the determination image generation part 230 is provided.
[0183] The separation image generation part 233 captured by taking pictures with the camera 110 generated captured image data and performs image separation processing of generating first separated image data mainly containing a color component belonging to the first wavelength band and second separated image data mainly containing a color component belonging to the second wavelength band from the captured image data.
[0184] The first separation image data are intended as image data of a primarily from the first illumination elements 121 emitted illumination light are generated image formed, and the second separation image data are intended as image data of a by mainly from the second illumination elements 131 emitted illuminating light formed image are generated. In other words, the first separation image data is data to be generated as data corresponding to the first captured image data in the first aspect of configuration, and the second separation image data is data to be generated as data corresponding to the second captured image data.
[0185] For example, in the case that lighting configured to emit red light as the first lighting elements 121 is used and lighting configured to emit blue light as the second lighting elements 131 is used, the first separation image data is generated as data mainly containing the R component and the second separation image data is generated as data mainly containing the B component. In such a case, the camera generates 110 preferably capture image data in the RGB format.
[0186] However, a component due to another type of light such as external light may be included in each separation image data unless the determination of the presence of defects by the defect determination part 240 is affected.
[0187] figure 20 is a flowchart showing the flow of the defect inspection in the defect inspection apparatus 2000 image pickup processing performed according to the present configuration aspect. In figure 20 and the associated description is similar to the description above regarding figure 10, the honeycomb structural body 1 also referred to as a “workpiece” as a defect inspection target and becomes the face 1a as an inspection target surface of the honeycomb structural body 1 also referred to as a "workpiece face".
[0188] First, the workpiece is placed on the table T (step S41 ), and then the image pickup execution part 100 moved to a pickup location (step S42 ), which corresponds to the flow of the defect inspection device1000 corresponds, and therefore the description of details is omitted.
[0189] After the image pickup execution part 100 is positioned at the recording point, the initial value of k = 1 is set (step S43 ) and become the first lighting part 120 associated k-th first lighting element 121 and that to the second lighting part 130 associated k-th second lighting element 131 switched on at the same time (step S44 ). However, the irradiation directions of the first lighting element 121 and the second lighting element 131 , which are turned on simultaneously, in an identical vertical plane. Then the camera takes 110 captures an image of the workpiece in such a power-on state (step S45 ). Accordingly, captured image data is generated as a piece of image data represented in the RGB format, for example.
[0190] The k-th captured image data obtained by the image capture is controlled by the image capture control part 111 to the separation image generation part 233 forwarded (step S46 ). After completion of the image recording and the forwarding, the k-th first lighting element 121 and second lighting element 131 , which are on, off (step S47 ). Alternatively, the k-th first lighting element 121 and second lighting element 131 be turned off immediately after the image acquisition is complete. Alternatively, all pieces of captured image data after completion of image capturing, which is performed during the m-th first lighting element 121 and second lighting element 131 are turned on to the separation image generating part 233 to get redirected.
[0191] If k = m does not hold (NO in step S48 ), in other words if there is any first lighting element 121 and any second lighting element 131 are still to be switched on, k = k + 1 is set (step S49 ) and become step S44 and repeats the following processing.
[0192] If k = m (YES in step S48 ), in other words, if all of the m first lighting elements 121 and second lighting elements 131 be switched on one after the other and the image recording by means of the camera 110 is then performed in each power-on state, the image pickup processing ends.
[0193] As a result, the image capture by the camera 110 , if each pair from the first lighting element 121 and the second lighting element 131 is successively turned on is then performed and m pieces of captured image data are obtained.
[0194] figure 21 is a flowchart showing the flow of the in the defect inspection apparatus 2000 color separation and reconstruction processing performed according to the present configuration aspect prior to the defect detection processing.
[0195] In the present configuration aspect, each of the m pieces of captured image data is obtained in a state that two illumination light beams having different wavelengths and irradiation angles from each other are irradiated in a superimposed manner, as described above. The color separation and reconstruction processing is processing of generating the first minimum luminance image data and the second minimum luminance image data based on the m pieces of captured image data similar to that in the first aspect of configuration.
[0196] First, the separation image generation part leads 233 performs color separation to sequentially attach to the m pieces of from the image pickup control part 111 to generate two pieces of separation image data forwarded to the recorded image data (step S51 ).
[0197] For example, it is in the case that the first lighting elements 121 emit red light, the second lighting elements 131emit blue light and the captured image data is generated in the RGB format, it is preferable that the first separation image data is generated by extracting only the R component from the captured image data to generate image data (R image data) and converting the R image data into luminance data and the second separation image data is generated by extracting only the B component from the captured image data to generate image data (B image data) and converting the B image data into luminance data.
[0198] However, the image data used to generate the first separation image data may also contain a color component other than the R component and the image data used to generate the second separation image data may also contain a color component other than the B component if the determination by the defect determination part 240 is not affected, for example when the pixel value is less than or equal to a predetermined threshold. In addition, a component due to outside light may be superimposed.
[0199] Those in the separation image generation part 233 generated first separation image data and second separation image data practically correspond by performing image pickup by the camera 110 , while only a single first lighting element 121 is switched on, obtained first recorded image data or by performing the image recording by means of the camera 110 , while only a single second lighting element 131 is turned on, obtained second recording image data.
[0200] When m pieces of the first separation image data and m pieces of the second separation image data due to the color separation by the separation image generating part 233 are generated, the minimum luminance imaging part generates 231 based on expression (1), first minimum luminance image data from the former (step S52 ) and generates second minimum luminance image data from the latter based on expression (2) (step S53 ).
[0201] The processing performed after obtaining the first minimum luminance image data and the second minimum luminance image data, specifically, the filter processing and the flaw detection processing is performed in the same manner as in the first aspect of configuration.
[0202] Thus, in the present configuration aspect, similarly to the first configuration aspect, a defect which needs to be detected can be reliably detected without misrecognizing any irregularity of a normal ceramic surface as a defect by detecting the existence of a defect consisting of a plurality of pieces minimum luminance image data generated from captured image data generated with different irradiation directions of the illumination light can be used.
[0203] Moreover, in the present configuration aspect, since illumination light from the first illumination element and illumination light from the second illumination element, which are separately irradiated in the first configuration aspect, are simultaneously irradiated, a time required for imaging is shortened compared to the first configuration aspect. Modification of the first embodiment
[0204] the figure 22 and figure 23 are drawings to show a modification of the arrangement and configurations of the first lighting part 120 and the second lighting part 130 .
[0205] In figure 8 is m = n = 8 and is the cut A1 - A1' a plane of symmetry of the first lighting elements 121a , the first lighting elements 121e , of the second lighting element 131a , and the second lighting element 131e , but this is in a case of the defect inspection device 1000 not strictly necessary according to the first configuration aspect. For example, as in figure 22 shows the arrangement of the first lighting elements 121 in the first lighting part 120 and the arrangement of the second lighting elements 131in the second lighting part 130 be set so that between the cut A2 - A2' as the plane of symmetry of the first lighting element 121a and the first lighting element 121e and the cut A3 - A3' as the plane of symmetry of the second lighting element 131a and the second lighting element 131e is a predetermined angle δ.
[0206] In figure 8 are the first lighting elements 121 and the second lighting elements 131 in the first lighting part 120 or the second lighting part 130 placed discreetly, but this is not essential. For example, as in figure 23, the first lighting part 120 and the second lighting part 130 also in a ring shape around the camera 110 spaced around and may be m equally spaced areas and n equally spaced areas (in figure 23 is m=n=8) the same as the first lighting elements 121 and the second lighting elements 131 be used.
[0207] Alternatively, the first lighting elements 121 or the second lighting elements 131 in one of the first lighting part 120 and the second lighting part 130 be discretely arranged as in figure 8 exemplifies, and they may be arranged in the other lighting part in a ring shape as in FIG figure 23 exemplifies.
[0208] At the in figure 8 are the first lighting elements 121 and the second lighting elements 131 are each arranged in a plane, but they do not have to be in an identical plane as long as the irradiation directions are arranged at equal angular intervals and the illuminance of the illumination light from the individual first illumination elements 121 and second lighting elements 131 is the same at one irradiation angle and one recording site.
[0209] In the first configuration aspect, when the second lighting elements are switched on 131 Image recording carried out after the first lighting elements are switched on 121 carried out image recording carried out, but this order can also be reversed. In this case, the second determination processing can be performed earlier than the first determination processing.
[0210] It may, depending on the structure of the honeycomb structural body 1 , a part called a connecting part in a lattice pattern on the face 1a exist, and in some cases there may be, depending on the material of the honeycomb structural body 1 (for example, when the material is SiC), give a bright part (part visually recognized as white in an image) corresponding to the connection part in an evaluation image. If the honeycomb structural body 1 consists of a plurality of parts (unit bodies) called honeycomb segments, the connecting part is a place where the honeycomb segments are connected to each other.
[0211] In the embodiment described above, a present one, the first honeycomb 3a in the detection image corresponding rectangular dark part excluded from the targets of detection processing by the opening mask processing, but when the bright part is present in the detection image, the bright part can be excluded from the targets of detection processing by the same mask processing in addition to the exclusion of the dark part by the opening mask processing. The array position and the size of the on the face 1a Namely, the present connection part are known, and thus the filter processing part 232, similarly to the opening mask processing, exclude the bright part corresponding to the connection part from the detection image by performing, for example, pattern matching, honeycomb expansion / contraction processing, and binarization processing using a maximum luminance value based on this information inputted in advance. image.
[0212] In the embodiment described above, the detection target is as a concave part on the face 1a of the honeycomb structural body 1 present defect, but the defect inspection method in the above-described embodiment is also applicable to a case where the detection target is a convex part on the face 1a of the honeycomb structural body 1 present defect (e.g. a protrusion).
[0213] Specifically, in the embodiment described above, the determination image generation part generates 230 first (second) minimum luminance image data in which a minimum value of luminance values of a plurality of pieces of first (second) pickup image data or a plurality of pieces of first (second) separation image data is set at an identical pixel position as a luminance value at the pixel position, and generates first (second) determination image data based on the first (second) minimum luminance image data, but instead it also generates first (second) maximum luminance image data in which a maximum value of luminance values of a plurality of pieces of first (second) captured image data or a plurality of pieces of first (second) separation image data at an identical pixel position is set as a luminance value at the pixel position, and generates first (second) evaluation image data based on the first (second) maximum luminance image data. In this case, the first (second) maximum luminance image data represents an image obtained by practically superimposing areas where the luminance value in the pieces of first (second) captured image data is high. Accordingly, the first (second) maximum luminance image data is data in which an area having a high luminance value due to a convex defect is enhanced.
[0214] Then, when there is a bright pixel area in an area greater than or equal to a predetermined bright part threshold in the detection image represented by both the first detection image data and the second detection image data, the defect detection part detects 240 that there is a convex defect at the occurrence position of the bright pixel area. Accordingly, a convex defect present on the outer surface of a ceramic body, which needs to be detected, can be reliably detected without misrecognizing an irregularity of a normal ceramic surface as a defect.
[0215] In the defect detection method of the embodiment described above, a plurality of pieces of captured image data in which a shadow area is variously formed are acquired by performing image capturing when each of a plurality of illumination elements configured to emit illumination light in irradiation directions different from each other , is turned on, and minimum luminance image data is generated based on the plurality of pieces of captured image data. Thus, multiple times of imaging is required to obtain a piece of minimum luminance image data, but imaging to obtain a piece of minimum luminance image data may vary depending on the configurations of the first lighting part 120 and the second lighting part 130 only be carried out once.
[0216] For example, the wavelength bands of the m are first illumination elements 121 emitted illuminating light are set to be different from each other, the wavelength bands of the n second illuminating elements are set 131 emitted illuminating light is set to be different from one another, the first image recording is carried out while the m first illuminating elements 121are switched on at the same time, and the second image recording is then carried out while the n second lighting elements 131 are switched on at the same time. Although each image pickup is performed while a plurality of illumination light beams are irradiated in a superimposed manner, the wavelength band and the irradiation direction are different from illumination light beam to illumination light beam. Thus, a plurality of pieces of separation image data generated by color separating the above obtained two pieces of captured image data into (m or n) pieces for each illumination light wavelength band practically corresponds to a plurality of pieces of captured image data in the first configuration aspect or a plurality of pieces of separation image data in the second configuration aspect. In other words, the generated plurality of pieces of separation image data can be subjected to the same filter processing and flaw detection processing as those in the first aspect of configuration described above. In such a case, the presence of a defect can only be determined by taking an image twice.
[0217] Also, in the case where n=m is set as in the second aspect of configuration, and the wavelength bands can be selected from the m first illumination elements 121 and the m second lighting elements 131 emitted illuminating light are set differently from one another, only one image recording can be performed while all the first illuminating elements 121 and all second lighting elements 131 are turned on at the same time, and captured image data obtained by image capture can be subjected to color separation for each illumination light wavelength band, in other words, into 2m pieces of separation image data, thereby determining the presence of a defect. Second embodiment Measures for enlarging the inspection target area
[0218] The method of defect inspection using the defect inspection apparatus described in the first embodiment 1000 or 2000 is not fundamentally subject to any limitation on the size of the honeycomb structural body 1 as an investigation target. Namely, the defect inspection can be performed on the honeycomb structural body by the method according to the first embodiment 1 of any size can be performed as long as the image pickup execution part 100 (especially the first lighting part 120 and the second lighting part 130 ) of the defect inspection device 1000 or 2000 is configured to suit the size of the honeycomb structural body 1 suitable. Alternatively, the defect inspection even if the inspection target area is opposite the size of the face 1a of the honeycomb structural body 1 is small by repeating the defect inspection many times while moving the inspection target area on the entire face 1a be performed.
[0219] In such a case, the number of repetitions of the inspection is one honeycomb structural body 1 smaller and the examination time is shorter because the examination target area is larger, especially when the angle of view (capturing range) of the camera 110 is enlarged. In practice, however, problems as described below become significant when the defect inspection is performed by the method according to the first embodiment with the inspection target area expanded. Problem due to the distance between the illuminating element and the irradiated position
[0220] figure 24 is a diagram showing a near the outer wall 2 of the honeycomb structural body 1 when irradiating the front 1a of the honeycomb structural body 1 image captured with illumination light Ls at a predetermined angle in an oblique direction (from the left side in the figure). IM7 indicates. The honeycomb body 1 includes a connection part forming a lattice pattern 7 at the front 1a . Then, in the in figure24 case shown, one agrees with the end face 1a parallel component (horizontal component) among the incident directions (optical axis directions) of the illumination light Ls with the extension direction of a connection part 7 at the front 1a match.
[0221] In the captured image IM7 a position closer to the illumination light source is photographed further to the left, and a position further away from the illumination light source is photographed further to the right. The illuminance of the illuminating light is inversely proportional to the square of the distance from a light source, and consequently there occurs a luminance difference between two locations whose distances from the illuminating light in the captured image IM7 are different even if there is no difference in surface condition between the two locations.
[0222] For example, both are from two points P1 and P2 , which are separated from each other along the horizontal component of the illumination light Ls as in FIG figure 24, points in the connection part 7 , and consequently the luminance between the two points is expected to be constant, but in practice there occurs a luminance difference Δb = b2 - b1 between the former luminance b1 and the latter luminance b2 .
[0223] When the inspection target area (capturing range) is expanded to reduce the number of repetitions of inspection and the inspection time, the influence of the luminance difference on the defect inspection processing cannot be neglected. Thus, when the inspection target area is expanded, measures against this point are required.
[0224] When the intersection between the optical axis L1 each lighting element and the front 1a with the intersection P between the optical axis CX of the camera 110 and the front side 1a coincides with the defect inspection apparatus used in the first embodiment 1000 or 2000 (please refer figure 9), as far as the vicinity of the intersection point P is concerned, it can be assumed that the distance to each lighting element is practically identical. In other words, it can be assumed that the illuminance of each lighting element in the immediate vicinity is identical. Thus, in the case that the size of the honeycomb structural body 1 is not very large and the front side 1a within the vicinity, a luminance difference between captured images due to a difference in the distance to each lighting element can be practically ignored.
[0225] However, if the angle of view (the shooting range) of the camera 110 is expanded (especially when in the camera 110 a wider-angle lens is used) to widen the inspection target area, the influence of a difference in the distance from each illuminating element to an irradiated position on the luminance at the irradiated position in the vicinity of an end part of the camera's angle of view 110 in one through the camera 110 captured image significant.
[0226] figure 25 is a drawing for describing the influence of a difference in the distance from the illumination element to the irradiated position.
[0227] figure 25(a) illustrates a situation in which the end face 1a of the honeycomb structural body 1 both with the illumination light Ln as well as with the illuminating light lf in relation to the normal direction (direction into the drawing) of the end face 1a obliquely irradiated in mutually symmetrical directions. The illumination light Ln becomes a near to a specific part in the vicinity of any end (hereinafter referred to simply as the end part) fv1 in the angle of view (shooting range) of the camera 110 located lighting element radiated, and the illumination light lfwill be from a far from the end part fv1 remote lighting element emitted. Even though figure 25(a) shows the two illumination light beams together for description, in practice, the illumination light beams are not emitted simultaneously.
[0228] the figure 25(b) and figure 25(c) are sectional views showing situations when the vicinity of an end part is irradiated fv1 contained defect (hole) df8 with the illumination light Ln and the illumination light, respectively lf illustrate schematically.
[0229] When irradiated with the illuminating light Ln as in figure 25(b) shows part of the face 1a , in which the defect df8 not present, and most of the defect df8 a radiation area RE11a of the illumination light Ln and is part of an inclined surface of the defect df8 a shadow area RE12a . Correspondingly, when irradiated with the illumination light lf as in figure 25(c) shows part of the face 1a , in which the defect df8 not present, and most of the defect df8 a radiation area RE11b of the illumination light Lf and is a part of a slanting surface of the defect df8 a shadow area RE12b .
[0230] When the illumination light Ln and the illumination light lf at the end part fv1 have the same illuminance is the luminance of two pieces by picture taking by the camera 110 shot image data obtained under irradiation with each illumination light in the two irradiation regions RE11a and RE11b is the same and is its luminance in the two shadow areas RE12a and RE12b the same. Furthermore, it can be said that the defect inspection method according to the first embodiment assumes that the luminance in at least the two irradiation regions RE11a and RE11b is the same.
[0231] However, when the illumination light Ln and the illumination light lf at the end part fv1 have an illuminance difference, which is due to a difference in the distance to the light source, the luminance, neither between the irradiation areas RE11a and RE11b still between the shadow areas RE12a and RE12b the same. In some cases, the luminance value is that of the illumination light Ln , for which the distance to the light source is small, formed shadow area RE12a greater than the luminance value of the illuminated light lf , for which the distance to the light source is large, formed irradiation area RE11b . For example, in some cases are the shadow areas RE12 not reflected in the investigation image. In such a case, it is difficult to carry out the defect inspection accurately. Measures such as balancing the luminance values of the irradiation areas are necessary to maintain the accuracy of the defect inspection RE11a and RE11b necessary. Thus, in the case that the inspection target area is expanded, measures against this point are required. Problem attributed to the relationship between the irradiation angle of the illumination light and the viewing angle
[0232] The influence of widening the angle of view (shooting range) of the camera 110 to widen the inspection target area also appears in the relationship between the irradiation angle of the illumination light and the angle of view. figure 26 is a drawing for describing the influence of the relationship between the irradiation angle of illumination light and the angle of view on a captured image.
[0233] figure 26(a) shows a case in which image pickup by the camera 110 is carried out while the end face 1a of the honeycomb structural body 1is irradiated obliquely with illumination light Lα at a predetermined angle (irradiation angle) θα with respect to the horizontal plane. In this case lies the first honeycomb 3a (hereinafter also referred to as a center part honeycomb 3a1) which has a rectangular opening just under a center part pos1 of a viewing angle fv2 the camera 110 and lies the first honeycomb 3a (hereinafter also referred to as an end part honeycomb 3a2) directly below an end part pos2 of the viewing angle fv2 . For convenience of description, the illuminance of the illumination light is Lα in the viewing angle fv2 constant.
[0234] When irradiated with the illumination light Lα in this way, a part of the end face becomes 1a , in which the first honeycomb 3a ( 3a1 , 3a2 ) not available, irradiation areas RE13a and RE14a of the illumination light Lα , which is also part of a wall surface 8 the first honeycomb 3a ( 3a1 , 3a2 ) near the front 1a contain. The larger the irradiation angle θα of the illumination light Lα, the larger that in the irradiation areas RE13a and RE14a contained area of the wall surface 8 .
[0235] Because the illumination light Lα the other part of the wall surface 8 and a deep part of the first comb 3a ( 3a1 , 3a2 ) is not reached, these parts become shadow areas RE13b and RE14b .
[0236] In other words, the irradiation state of the illumination light Lα is at the center portion honeycomb 3a1 and at the end part honeycomb 3a2 the same.
[0237] However, the middle section honeycomb exhibit 3a1 and the end part honeycomb 3a2 when forming an image in a when image pickup is performed by the camera 110 The captured image obtained in such an irradiation state has a difference as described below.
[0238] First is a picture of the center section honeycomb 3a1 a rectangular dark part SD13 with a luminance value of essentially zero.
[0239] However, a picture is of the end part honeycomb 3a2 mostly a dark part SD14a with a substantially zero luminance value, similar to the dark part SD13 , but it also contains a faint dark part SD14b with a luminance a little higher than that of the dark part SD14a (a little lighter enough to be visually recognizable) which has a predetermined width. This corresponds to the reflection in the irradiation area RE14a contained part of the wall surface 8 . The larger the irradiation angle θα, the clearer this reflection.
[0240] That is, a difference in the area of the dark part occurs between the center part honeycomb image 3a1 and the picture of the end part honeycomb 3a2 on. Thus, when the detection image (detection image data) is (are) generated based on a captured image (captured image data) containing such a reflection, there may be such an error that correct recognition is not performed , may occur although an entire dark part SD14 of the dark part SD14a and the faint dark part SD14b as one of the first honeycomb 3a associated honeycomb area must be recognized. figure 26(b) is an illustration showing such a determination image IM8 exemplified. In the investigative picture IM8 there is a faint dark part in part E as a reflection of the wall surface 8 before.
[0241] Such a mistake becomes significant when the viewing angle fv2 is enlarged to expand the inspection target area. Thus, this point requires measures when the inspection target area is expanded. defect inspection device
[0242] figure 27 is a block diagram showing the configuration of the defect inspection apparatus 3000 according to the present embodiment. Components of the defect inspection device 3000are largely the same as the components of the defect inspection apparatus 1000 according to the first embodiment, and hence all common parts are denoted by same reference numerals and detailed description thereof is omitted and differences from the defect inspection apparatus are mainly discussed hereinafter 1000 described.
[0243] As in figure 27 includes the defect inspection apparatus 3000 similar to the defect inspection device 1000 above all: the table T on which the honeycomb structural body 1 in a position in which the end face 1a is horizontal as the inspection target surface is to be laid; the image pickup execution part 100 , which is configured for the imaging of the face 1a by means of the camera 110 during frontal irradiation 1a of the honeycomb structural body placed on the table T 1 to perform with illuminating light; and the controller 200 , which is configured to control the image pickup executive 100 and the defect determination based on a by the image pickup execution part 100 obtained captured image.
[0244] However, the defect inspection device 3000 according to the present embodiment versus a case where the defect inspection device 1000 according to the first embodiment, be able to accurately perform defect inspection in a large inspection target area. In the present embodiment, the relationship is that the inspection target area of the inspection apparatus 3000 is larger than that of the inspection apparatus 1000 according to the first embodiment, from one on the assumption that the distance (camera distance) between the camera 110 and the front side 1a of the honeycomb structural body 1 as the inspection target surface is substantially the same in the devices or that the camera distance in the inspection device 3000 is smaller than that of the inspection device 1000 , based comparison derived. The reason is that in such a case, although the imaging target area increases as the camera distance in the inspection apparatus increases 1000 according to the first embodiment, problems such as increase in size of the inspection apparatus 1000 and decrease in image resolution, which is not desirable.
[0245] In order to cope with such an expansion of the examination target area, the camera 110 with lens specifications that enable imaging of a larger area than that in the defect inspection apparatus 1000 used camera 110 allow, used. The lens specifications of the actual camera 110 can be set based on an examination condition (such as the size of the examination area in a single examination).
[0246] figure 28 is a bottom view of a main part of the image pickup execution part 100 (with the image pickup execution part viewed from bottom to top in the vertical direction 100 ), and figure 29 is a sectional view along the line A4 - A4' in figure 28. The cut A4 - A4' in figure 28 is the optical axis CX of the camera 110 containing vertical section, is a plane of symmetry of a small-angle lighting element 116a , a small-angle lighting element 116e , a mid-angle lighting element 121a and a mid-angle lighting element 121e , to be described later, is one between a large-angle lighting element 131a and a large angle lighting element 131b running plane and is also one between a large-angle lighting element 131e and a large angle lighting element 131f running level. However, the cut runs A4 - A4' in figure 29 to facilitate illustration by the large-angle illuminating element131a and the large angle lighting element 131e .
[0247] In addition, shows figure 29 additionally the one on the table for a better understanding T , which in figure 29 is not shown laid honeycomb structural body 1 . the figure 28 and figure 9 contain right-hand xyz coordinates whose z-axis direction is along the vertical direction, whose x-axis direction is along the right-left direction in figure 28 and whose y-axis direction is along the up-down direction in the figure. Accordingly is figure 29 showing the cut A4 - A4' in figure 28 illustrates a zx sectional view.
[0248] The configurations and placement of the table T and the camera 110 in the defect inspection device 3000 according to the present embodiment are basically the same as those in FIG figure 8 and figure 9 exemplary defect inspection apparatus 1000 . Upon inspection, the honeycomb structural body 1 namely placed on the table T (not shown) so that the end face 1a as the inspection target surface is a horizontal top as in figure 29 shown. In the image pickup execution part 100 is the camera 110 provided in a position in which its lens faces vertically downward and is its optical axis CX aligned in the vertical direction to obtain a shooting target in the vertical-down direction. Accordingly, the camera can 110 the image acquisition in one at an intersection P between the optical axis CX and the front side 1a centered predetermined area.
[0249] In the image pickup execution part 100 the defect inspection device 1000 according to the first embodiment, the two lighting parts are the first lighting part 120 and the second lighting part 130 around the camera 110 arranged around but in the defect inspection device 3000 according to the present embodiment, three lighting parts are one small-angle lighting part 115 , a medium-angle lighting part 120 and a wide-angle lighting part 130 by an appropriate arranging device (not shown) on the underside of the image pickup execution part 100 included holding body 101 around the camera 110 arranged around. Similar to the defect inspection device 1000 is the camera 110 at least when recording the image in the holding body 101 intended opening 102 used and leaves the holding body 101 , on which the camera 110 and each lighting part are arranged, by means of the moving mechanism 140 move.
[0250] More specifically, the medium-angle lighting part correspond 120 and the wide-angle lighting part 130 the first lighting part 120 or the second lighting part 130 the defect inspection device 1000 according to the first embodiment. The following are components of the medium-angle lighting part 120 and the large-angle lighting part 130 with the same reference numerals as the corresponding components of the first lighting part for better understanding 120 and the second lighting part 130 designated.
[0251] That is, the medium-angle lighting part 120 has a configuration in which m1 ( m1 ≥ 4) Mid-angle lighting elements 121 of the same power, each having the irradiation angle θ1 (preferable is θ1 = 30° to 60°), in a horizontal plane at equal angular intervals around the camera 110 are arranged around. the figure 28 and figure 29 exemplify a case of m1 = 8. In other words, the figure 28 and figure 29 exemplify a case in which the eight middle-angle lighting elements 121 ( 121auntil 121h ) are provided. the figure 28 and figure 29 also show as each medium-angle lighting element 121 for example, bar lighting in which a large number of LED elements are lined up in a rectangular shape.
[0252] The wide angle lighting part 130 has a configuration in which m2 ( m2 ≥ 4) Large angle lighting elements 131 of the same power, each having the irradiation angle θ2 (preferable is θ2 = 60° to 85°), in a horizontal plane at equal angular intervals around the camera 110 are arranged around. However, those illustrate figure 28 and figure 29 more specifically exemplifies a case of m2 = 8, in which the large-angle lighting part 130 is provided as ring lighting in which a large number of LED elements are lined up concentrically in a ring shape as in FIG figure 23, and become areas obtained by equally dividing the ring illumination into eight parts as the respective large-angle illumination elements 131 ( 131a until 131h ) used.
[0253] In this way, in the figure 28 and figure 29 m1 = m2 = 8, but m1 = m2 is not strictly necessary and it can also m1 ≠ m2 are valid. In figure 28 are the arrangement positions of the individual medium-angle lighting elements 121 versus the arrangement positions of the individual large-angle lighting elements 131 in a direction around the camera 110 shifted around in the horizontal plane (circumferential direction) by 22.5°, but this is not essential and the arrangement positions of the individual middle-angle lighting elements 121 can also be related to the arrangement positions of the individual large-angle lighting elements 131 coincide in the circumferential direction as in figure 8 shown.
[0254] Moreover, as described above, the defect inspection device includes 3000 according to the present embodiment, the small-angle lighting part 115 in addition to the medium-angle lighting part 120 and the large-angle lighting part 130 .
[0255] the figure 28 and figure 29 exemplify a configuration in which the small-angle illumination part 115 eight low-angle lighting elements 116 ( 116a , 116b , 116c , 116d , 116e , 116f , 116g and 116h ) contains. In this case, they have eight low-angle lighting elements 116 each have a longitudinal direction along the horizontal direction and are arranged so as to have angular intervals of 45° from each other in a horizontal plane, and are each in an inclined position on the holding body 101 fastened like the one in figure 29-29 exemplified small-angle lighting elements 116a and 116e . Similar to the mid-angle lighting element 121 show the figure 28 and figure 29 than any low-angle lighting element 116 for example, bar lighting in which a large number of LED elements are lined up in a rectangular shape.
[0256] More specifically, has an irradiation direction D0 each low-angle lighting element 116 a predetermined angle θ0, which is smaller than the angle θ1 with respect to the horizontal plane, and the intersection between an optical axis falls L0 same and the front side 1a with the intersection P between the optical axis CX the camera 110 and the front side 1a together, similar to the optical axis L1 of the mid-angle lighting element 121 and the optical axis L2 of the large-angle lighting element. Accordingly, the eight small-angle lighting elements 116irradiate substantially identical areas centered on the intersection P with illumination light at the same irradiation angle θ0 in directions different from each other at angular intervals of 45° from each other in the horizontal plane.
[0257] The multitude of small-angle lighting elements 116 , the plurality of medium-angle lighting elements (first lighting elements) 121 and the plurality of large-angle lighting elements (second lighting elements) 131 are in different to the front side 1a separated by parallel planes as the examination target surface.
[0258] The irradiance angle θ0 of each small-angle lighting element 116 is preferably 5° to 30° and is, for example, 15°.
[0259] In the first embodiment, an irradiation angle is θ1 of the first lighting element 121 , which is less than 30°, is not desirable because a normal surface irregularity ns of the ceramic surface 6 shadow area occurring becomes larger and the surface irregularity ns is erroneously recognized as a defect with a high probability. However, in the present embodiment, while the irradiation angle is θ1 for the first lighting elements 121 corresponding mid-angle lighting elements 121 is maintained, the small-angle lighting part 115 with the irradiation angle θ0, which is smaller than θ1, separate from the middle-angle lighting part 120 intended to enlarge the survey target area. As will be described later, FIG. 1 works by taking an image using the small-angle lighting part 115 detection image data obtained using detection image data obtained, having a threshold other than that set by image pickup using the medium-angle lighting part 120 determination processing using determination image data to be traced back to captured image data obtained. Thus, the problem of false detection described above does not occur.
[0260] The number in the small-angle lighting part 115 provided small-angle lighting elements 116 is not limited to eight, but typically can m0 ( m0 ≥ 4) Small-angle lighting elements 116 of the same power, each having the irradiation angle θ0, in the horizontal plane at equal angular distances from each other around the camera 110 be provided around arranged positions.
[0261] In figure 28 shows the arrangement positions of the individual small-angle lighting elements 116 with the arrangement positions of the individual medium-angle lighting elements 121 in the circumferential direction in the horizontal plane, but this is not essential, and their arrangement positions may also be shifted from each other like the relationship between the arrangement positions of the medium-angle lighting elements 121 and the arrangement positions of the large-angle lighting elements 131 .
[0262] Similar to the defect inspection device 1000 according to the first embodiment is obtained in the defect inspection apparatus 3000 according to the present embodiment m0 Pieces of captured image data (small-angle captured image data) by subsequent image capture by the camera 110 every time the m0 (for example, eight) in the small-angle lighting part 115 provided small-angle lighting elements 116 be switched on one after the other. You get accordingly m1 Pieces of captured image data (middle-angle captured image data) by subsequent image capture each time the m1 (e.g. eight) in the medium-angle lighting part 120 provided mid-angle lighting elements 121 be switched on one after the other. In addition, you get m2 Pieces of captured image data (wide-angle captured image data) by subsequent image capture each time the m2(for example, eight) in the large-angle lighting part 130 provided large-angle lighting elements 131 be switched on one after the other.
[0263] One of the reasons why the defect inspection device 3000 the small-angle lighting part 115 contains is an error due to a reflection of the wall surface 8 the first honeycomb 3a in a captured image due to the expansion of the inspection target area as described above. figure 30 is a drawing for describing the relationship between that of the small-angle lighting part 115 emitted illuminating light and the viewing angle.
[0264] figure 30 shows a case in which image pickup by the camera 110 is carried out while the end face 1a of the honeycomb structural body 1 with illumination light Lβ from a small-angle illumination element 116 is irradiated obliquely at a predetermined angle (irradiation angle) θ0 = θβ (< θα) with respect to the horizontal plane. It is assumed that the viewing angle fv2 the camera 110 and the arrangement positions of the center-part honeycomb 3a1 and the end part honeycomb 3a2 in this case the same as those in the in figure 26(a) are.
[0265] With such irradiation with the illumination light Lβ become part of the face 1a , in which the first honeycomb 3a ( 3a1 , 3a2 ) not available, irradiation areas RE15a and RE16a of the illumination light Lβ , which is also part of the wall surface 8 the first honeycomb 3a ( 3a1 , 3a2 ) near the front 1a contain. From the other part of the wall surface 8 and a deep part of the first comb 3a ( 3a1 , 3a2 ) become shadow areas RE15b and RE16b .
[0266] However, in this case the figure 26(a) an irradiation angle θβ of the illumination light Lβ smaller than the irradiation angle θα of the illumination light Lα , and consequently the area of in the irradiation areas RE13a and RE14a contained wall surface 8 smaller than that in the case of irradiation with the illumination light Lα .
[0267] The appearance of images of the center section honeycomb 3a1 and the end part honeycomb 3a2 in a picture taking by means of the camera 110 captured image obtained in such an irradiation state is compared with that in the case of FIG figure 26(a) as described below.
[0268] First is the picture of the center section honeycomb 3a1 similar to the case of figure 26(a) a rectangular dark part SD15 with a luminance value of essentially zero.
[0269] The image of the end part honeycomb 3a2 is mostly a dark part SD16a with a substantially zero luminance value like the dark part SD15 . Similar to the case of figure 26(a) becomes a reflection of the wall surface 8 due faint dark part SD16b formed, but the area of formation is sufficiently small compared to the dark part SD16a and negligible in its effect on examination accuracy.
[0270] This means that evaluation image data with which a honeycomb area can be recognized correctly can be obtained by performing image pickup using the small-angle illumination part 115 can get.
[0271] In addition to such provision of the small-angle lighting part 115 are the upper and lower halves of each lighting element each in the imaging execution part 100 provided lighting part in the defect inspection device 3000 individually dimmable according to the present embodiment.
[0272] Special are the top and bottom halves of each small-angle lighting element 116 a dimming unit 116Uor a dimming unit 116L , which are individually dimmable. The light quantities of the dimming unit are special 116U and the dimming unit 116L individually adjustable. Correspondingly, an upper dimming unit 121U and a lower dimming unit 121L each mid-angle lighting element 121 individually dimmable. Also include a top dimming unit 131U and a lower dimming unit 131L each high-angle lighting element 131 individually dimmable.
[0273] Thus, the small-angle lighting elements 116 ( 116a until 116h ), the mid-angle lighting elements 121 ( 121a until 121h ) and the high-angle lighting elements 131 ( 131a until 131h ) each arranged in its entirety so that the respective optical axes L0 , L1 and L2 through the intersection P between the optical axis CX the camera 110 and the front side 1a of the honeycomb structural body 1 run, but the optical axis of each dimming unit is shifted from the intersection point P. The optical axes of the dimming units run in a special way 116L , 121L and 131L through the front of the intersection point P and pass the optical axes of the dimming units 116U , 121U and 131U through the back of the intersection P .
[0274] The dimming of each dimming unit is under the control of the lighting control part 220 . An LED element having a principal ray angle FWHM of about 5° to 30° (for example, 12° when the distance from each lighting element to the intersection point P is about 180 mm) is preferably used to perform dimming individually and excellently. However, at a long distance from the illumination to the intersection point P, the principal ray angle FWHM is preferably small because the illumination light diverges before reaching an examination object. With a small distance from the illumination to the point of intersection P, the chief ray angle FWHM is preferably large.
[0275] figure 31 is a diagram for describing the effect of dimming (individual dimming) of each dimming unit. is special figure 31 is a graph showing the relationship between the horizontal distance from the illumination (light source) and the luminance (luminance distribution) when image pickup of a uniform flat surface is performed while the flat surface is illuminated with illumination light from a small-angle illumination element 116 is irradiated in an oblique direction.
[0276] As described above, the illuminance of the illuminating light is inversely proportional to the square of the distance from a light source. Consequently, with increasing horizontal distance from the lighting (light source), the luminance decreases when individual dimming of the dimming units 116L and 116U is not carried out, monotonously as with "without dimming" in figure 31 shown. In the case of "without dimming" in figure 31 there is a difference in luminance Δb1 at both ends of the shooting range (angle of view). The same also applies in the event that a lighting element can only be dimmed in its entirety and individual dimming cannot be carried out.
[0277] On the other hand shows figure 31 at "with dimming" an example in which individual dimming by the dimming units 116L and 116U is performed to thereby increase the luminance on a side far from the illumination than that in the case of "without dimming", while the luminance on a side close to the illumination is substantially the same as that in the case of " without dimming”. Specifically, dimming is performed to reduce the amount of light from the upper dimming unit 116U each low-angle lighting element 116 compared to the amount of light from the lower dimming unit 116L to increase relatively.
[0278] In such a case, the luminance between the side near the illumination and the center in the shooting range is substantially constant or a little higher closer to the center, and the luminance difference Δb2 at both ends of the shooting range is smaller than the luminance difference Δb1 in the case of " without dimming".
[0279] In the defect inspection device 3000 According to the present embodiment, a luminance difference corresponding to the difference in the distance of each lighting element in the shooting range can be reduced by performing such individual dimming in advance before the inspection for all of the small-angle lighting elements 116 , the mid-angle lighting elements 121 and the high-angle lighting elements 131 is carried out.
[0280] Specific methods and requirements of the individual dimming are not particularly limited, but, for example, criteria for the lowest luminance or the luminance difference Δb2 are given, and dimming is performed to meet the criteria.
[0281] Instead of dimming based on the luminance distribution in a captured image, the illuminance in the capture area can be directly measured by predetermined measuring devices, and individual dimming can be performed based on its distribution (illuminance distribution).
[0282] The following is the controller 200 described. In the defect inspection device 3000 according to the present embodiment is similar to the defect inspection apparatus 1000 according to the first embodiment, the controller 200 realized by a computer such as a general-purpose personal computer. The control 200 mainly contains the input control part 201 and the display part 202 and also includes functional components realized by execution of an operating program stored in a storage part (not shown) such as a hard disk provided in the computer by a control part (not shown) provided in the computer including CPU, ROM and RAM.
[0283] However, a comparison between figure 7 and figure 27 clearly shows that the controller 200 the defect inspection device 3000 according to the present embodiment differs therein from the controller 200 the defect inspection device 1000 according to the first embodiment differs in that, in addition, a luminance correction processing part 234 in the determination image generation part 230 is provided.
[0284] The luminance correction processing part 234 captured by taking pictures with the camera 110 generated captured image data (small-angle captured image data, middle-angle captured image data, large-angle captured image data), and performs luminance correction processing of correcting the luminance distribution of the captured image data.
[0285] Usually, the luminance correction processing is in the luminance correction processing part 234 a processing performed to obtain the luminance value at the face 1a of the honeycomb structural body 1 between the pieces of small-angle pickup image data, between the pieces of medium-angle image data and between the pieces of large-angle image data to compensate for the generation of an error due to the difference in the distance from the illumination (light source) such as based on figure 25 before minimum luminance image data ( m0 pieces of small-angle minimum luminance image data, m1 Pieces of mean-angle minimum luminance image data and m2 pieces of high-angle minimum luminance image data). m0 pieces of small-angle photographed image data, m1 Pieces of mid-angle photographed image data or m2pieces of large-angle captured image data are obtained to generate detection image data (small-angle detection image data, medium-angle detection image data, and large-angle detection image data). The first honeycomb is opened schematically in this processing 3a , the connecting part 7 or a normal part where there is no defect or the like on the front side 1a is set as a reference part (base part), and the luminance is balanced at the reference part between the pieces of captured image data.
[0286] The luminance correction processing also has an effect of reducing the luminance difference due to a difference in the distance from the illumination light in the viewing angle, which persists even after the above-described individual dimming of each dimming unit.
[0287] In the defect inspection device 3000 according to the present embodiment, on the basis of each piece of captured image data, by means of the luminance correction processing by the luminance correction processing part 234 generated corrected image data ( m0 pieces of corrected small-angle image data, m1 Pieces of corrected mid-angle image data and m2 pieces of corrected large-angle image data) for generating minimum luminance image data (small-angle minimum luminance image data, medium-angle minimum luminance image data, large-angle minimum luminance image data) by the minimum luminance image generation part 231 used.
[0288] Details of the luminance correction processing by the luminance correction processing part 234 will be described later. defect investigation processing
[0289] The following description of the through the defect inspection device 3000 The defect inspection processing performed mainly deals with differences from the defect inspection device 1000 according to the first embodiment. image capture processing
[0290] figure 32 is a flowchart showing a schematic flow of the defect inspection with the defect inspection apparatus 3000 performed image pickup processing.
[0291] In the image pickup processing, first, similarly to the first embodiment, a workpiece (honeycomb structural body 1 as a defect inspection target) in a posture where a front side thereof is a top side on the table T laid (step S61 ) and will after placing if via the input panel 201 an execution instruction for a defect inspection is given, the movement mechanism 140 driven to the image pickup execution part 100 (specifically the camera 110 and the small-angle lighting part 115 , the mid-angle lighting part 120 and the large-angle lighting part 130 supporting holding body 101 ) to move to a pick-up location (step S62 ).
[0292] In this case, the workpiece, similar to the first embodiment, when placed on the table T is laid, so can be positioned or the position of the camera 110 be adjusted in the horizontal plane so that the honeycomb 3 of the workpiece (the first honeycombs 3a in appearance) in the shooting range of the camera 110 , which is defined in a rectangular shape, are lined up in the long axis and short axis directions. However, even if the row direction of the honeycomb 3 with respect to the longitudinal and transverse axis directions in the shooting range of the camera 110 is a little skewed, the detection processing can be performed smoothly by making a correction in the detection processing considering the skewing as needed.
[0293] Similar to the first embodiment, a sensor configured to sense whether the workpiece is placed on the table T may be provided and the integrated control part 210in response to a sensing signal from the sensor, a predetermined control signal for sequentially executing the image pickup processing and the subsequent detection processing to each component of the defect inspection apparatus 3000 send.
[0294] After the image pickup execution part 100 is located at the pickup position, the image pickup is performed using the small-angle illumination part 115 (Step S63 ), the image acquisition using the medium-angle lighting part 120 (Step S64 ) and the image acquisition using the large-angle illumination part (step S130 ) performed one after the other. As described above, such image pickup is performed after individually dimming each lighting element is performed in advance to reduce the luminance difference in the pickup area.
[0295] figure 33 is a flowchart showing a specific flow of each type of image pickup processing. The flow of each type of image pickup processing is the same as the flows of image pickup processing using the first lighting part 120 and the image pickup processing using the second lighting part 130 in the first embodiment, which in figure 10 are shown individually.
[0296] Namely, in any kind of image pickup processing, the initial value of p = 1 is set (step S71 ) and the consecutive image recording by the camera 110 carried out while all lighting elements are switched on one after the other.
[0297] Specifically, this becomes each lighting part (the small-angle lighting part 115 , the mid-angle lighting part 120 or the wide-angle lighting part 130 ) associated p-th lighting element (small-angle lighting element 116 , medium-angle lighting element 121 or high-angle lighting element 131 ) switched on (step S72 ) and takes the camera 110 captures an image of the workpiece in such a power-on state (step S73 ). The p-th captured image data (small-angle captured image data, middle-angle captured image data, or large-angle captured image data) obtained by the image capture is controlled by the image capture control part 111 to the luminance correction processing part 234 forwarded (step S74 ) and subjected to detection processing to be described later. After completion of the image acquisition and the forwarding, the switched-on p-th lighting element (small-angle lighting element 116 , medium-angle lighting element 121 or high-angle lighting element 131 ) switched off (step S75 ). Alternatively, the pth lighting element can be switched off immediately after the image recording is complete. Alternatively, each piece of captured image data (small-angle captured image data, medium-angle captured image data, or large-angle captured image data) can be sent to the minimum luminance imaging part 231 be forwarded when all lighting elements of each lighting part are used for image acquisition and the last image acquisition is completed.
[0298] If not all lighting elements are used (NO in step S76 ), in other words, if there is any lighting element that has yet to be turned on, p = p + 1 is set (step S77 ) and become step S72 and repeats the following processing.
[0299] If all lighting elements are used (YES in step S76 ), the image pickup processing using the lighting part ends. Luminance correction processing
[0300] figure 34 is a flowchart showing a schematic flow of the defect inspection with the defect inspection apparatus 3000 by means of the luminance correction processing 334 performed luminance correction processing (small-angle correction processing, medium-angle correction processing and large-angle correction processing).figure 35 is a diagram exemplifying the processing content of the luminance correction processing.
[0301] In this description, it is assumed that recorded image data has a luminance distribution pf1 as in figure 35(a). Specifically, represents a pixel area RE21 with a luminance value significantly lower than that of its surroundings, an image of the first comb 3a as an opening, represents a pixel area RE22 with a luminance value that is significantly higher than that of its surroundings, an image of the connection part 7 represents and represents a pixel area RE23 with a luminance value slightly lower than that of its surroundings, an image of one at the front 1a formed defect (typically a hole). In the following, a part other than these pixel areas RE21 , RE22 and RE23 referred to as a base part. The negative slope of the luminance distribution pf1 in the figure in its entirety including the base part indicates the luminance difference remaining even after individual dimming. The connection part 7 exists depending on the configuration of the honeycomb structural body 1 may not be the target of investigation.
[0302] In the luminance correction processing, first, an average value (average luminance value) Avr of the luminance value at each pixel for captured image data having such a luminance distribution pf1 have, calculated (step S81 ). In figure 35(a), the average luminance value Avr is shown as a broken line.
[0303] After obtaining the average luminance value Avr, the luminance value of the first comb 3a like the pixel area RE21 and the luminance value of the connection part 7 like the pixel area RE22 , if it exists, is replaced by the average luminance value Avr (step S82 ). figure 35(b) shows a luminance distribution pf2 of image data (data after replacement) after replacement, and also shows the luminance distribution pf1 as a dashed line. The arrangement positions and sizes of the first honeycomb 3a and the connecting part 7 are known and thus the positions of pixels of their images can be specified and the replacement can be easily performed.
[0304] Then, after performing the replacement, smoothing processing is performed on the data after replacement to generate smoothing data (step S83 ). A well-known method is applicable to the smoothing processing. figure 35(c) shows a luminance distribution pf3 of the smoothing data obtained.
[0305] The luminance distribution provided by the smoothing data pf3 has a negative slope in the figure like the luminance distribution provided by the captured image data subjected to the luminance correction processing pf1 . In other words, the luminance distribution provided by the smoothing data obtained by preliminarily generating the post-replacement data and smoothing the post-replacement data pf3 shows a rough distribution tendency of the luminance of the other part except parts like the first comb 3a and the connecting part 7 , which are known singular points in the luminance distribution pf1 are.
[0306] The luminance value in the smoothing data derived from captured image data for which the distance from a lighting element used for imaging is small tends to be larger in its entirety than the luminance value in the smoothing data derived from captured image data for which the distance from the lighting element is large smoothing data.
[0307] After obtaining the smoothing data, the difference between the original recording image data showing the luminance distribution pf1 provide, and the smoothing data is generated as corrected image data (step S84 ). figure 35(d) shows a luminance distribution pf4of the corrected image data obtained. More specifically, the corrected image data is obtained by calculating the difference in luminance values at an identical pixel position between the captured image data and the smoothing data for all pixel positions.
[0308] As in figure 35(d) contains the luminance distribution represented by the corrected image data pf4 , similar to those in figure 35(a), that of the first honeycomb 3a corresponding pixel area RE21 , which the connecting part 7 corresponding pixel area RE22 and one on the front 1a pixel area corresponding to the formed defect (typically a hole). RE23 . The luminance value in the base part other than these pixel areas is substantially constant. This is the effect of subtracting the luminance value of the smoothing data, which in the figure has more of a negative slope than the original captured image data, from the original captured image data.
[0309] Since the luminance value of the base part is essentially constant in this way, the difference in luminance values due to the distance from the lighting element used for imaging is removed from the corrected image data.
[0310] In addition, since the smoothing data of the luminance value corresponding to the distance from the illuminant is subtracted from the original photographed image data, the luminance of the base part, which is a normal (defect-free) part of the end face 1a is and through all lighting elements in each of the small-angle lighting part 115 , the mid-angle lighting part 120 and the large-angle lighting part 130 is irradiated in the same way as are considered equal (substantially identical) between pieces of corrected small-angle image data, between pieces of corrected medium-angle image data and between pieces of corrected large-angle image data. Accordingly, the luminance value difference due to the difference between the distances between different lighting elements is also eliminated. discovery processing
[0311] The defect inspection device 3000 according to the present embodiment performs small-angle determination processing based on m0 pieces of corrected small-angle image data obtained by the luminance correction processing as described above, small-angle detection processing based on m1 Pieces of corrected medium-angle image data and small-angle detection processing based on m2 pieces of corrected mean-angle image data.
[0312] First, each piece of corrected image data (corrected small-angle image data, corrected medium-angle image data, and corrected large-angle image data) is sent to the minimum luminance image generation part 231 subjected to and used to generate minimum luminance image data (low-angle minimum luminance image data, medium-angle minimum luminance image data, large-angle minimum luminance image data) (see figure 27). The generation of the corrected image data is similar to the generation of the first minimum luminance image data in step S21 the in figure 11 and the generation of the second minimum luminance image data in step S31 the in figure 14 second determination processing.
[0313] Then, by means of the minimum luminance imaging part 231 generated minimum luminance image data of filter processing by the filter processing part 232 subjected. Determination image data (small-angle determination image data, medium-angle determination image data and large-angle determination image data) are generated by means of the filter processing (see figure 27). The filter processing is the same as the filter processing in step S22 the in figure 11 shown first determination processing and step S32 the in figure 14 second determination processing.
[0314] Then, using the obtained determination image data, the processing for determining the presence of defects is performed by the defect determination part 240 carried out (see figure 27).
[0315] figure 36 is a flowchart showing the flow of the processing performed by the defect determination part 240 investigation processing performed. As in figure 36 is used in the defect inspection apparatus 3000 according to the present embodiment, the determination (steps S91 until S93 ) based on the through the filter processing part 232 generated small-angle detection image data, medium-angle detection image data and large-angle detection image data are performed stepwise. Each type of detection processing is referred to as small-angle detection processing, medium-angle detection processing, and large-angle detection processing.
[0316] If the determination image has a connection part 7 corresponding bright part, the opening mask processing is also performed on the bright part as in the modification of the first embodiment described above.
[0317] If the inspection target is the vicinity of the outer wall 2 of the honeycomb structural body 1 is as in the picture taken IM7 in figure 24 is an image of an external space (workpiece background) EX of the honeycomb structural body 1 corresponding part is a dark part in an examination pickup image, and hence the aperture mask processing is also performed on the dark part corresponding to the external space EX. The same applies when a captured image including a dark part corresponding to such an outer space is subjected to the inspection in the first embodiment.
[0318] From a comparison between figure 26(a) and figure 30 shows that when the irradiation angle of the illumination light is small, the arrangement position and the size of an image of one of the first honeycomb 3a corresponding rectangular dark part (comb area) in a captured image probably with the arrangement position and the size of the actual first comb 3a coincide at any position in the viewing angle. This can be used so that the pattern is one of the first honeycomb 3a corresponding dark part described in the small-angle minimum luminance image data is used as a mask pattern in the aperture mask processing.
[0319] The flow of the specific processing of each detection processing is the same as that of the processing in step S23 and afterwards in figure 11 and the processing in step S23 and afterwards in figure 14. Specifically, the content of the mean angle determination processing is identical to that of FIG figure 11, and the content of the large-angle detection processing is identical to that of FIG figure 14 second determination processing.
[0320] The execution flow of the small-angle determination processing performed before these processes is the same as those of the first determination processing (middle-angle determination processing) and the second determination processing (large-angle determination processing). However, in the small-angle detection processing, one for the steps S23 and S33 corresponding determination used threshold (first dark part threshold) is set to a larger value than the first threshold in the first determination processing (mean angle determination processing) and becomes one for the steps S27 and S37 corresponding determination used threshold (second dark part threshold) is set to a larger value than the second threshold in the first determination processing (mean angle determination processing).
[0321] This is due to the expansion of the camera's angle of view 110to expand the study target area. Specifically, this takes into account the possibility that in the middle-angle determination processing based on using the middle-angle lighting part 120 obtained mid-angle photographed image data, a pinhole / chipping detection, which is understood to be detection processing identical thereto in the first detection processing in the first embodiment, cannot be sufficiently performed.
[0322] In other words, when the relatively large honeycomb structural body 1 in the defect inspection device 3000 according to the present embodiment, a defect such as a hole or a chip with a relatively small depth is detected by the small angle detection processing, a crack with a large depth is detected by the large angle detection processing and becomes a defect with an intermediate depth is detected by the mean angle detection processing.
[0323] Similar to the first embodiment, the presence of a defect is comprehensively determined based on detection processing results of the small-angle detection processing, the medium-angle detection processing, and the large-angle detection processing (step S94 ). A result of the determination is used as determination result data by the defect determination part, as necessary 240 to the integrated control part 210 Posted. When it is determined by at least one of the small-angle determination processing, the medium-angle determination processing, and the large-angle determination processing that there is a defect at any point of the inspection target area, the defect determination part determines 240 that there is a defect in the inspection target area.
[0324] The integrated control part 210 controls the display part based on the description content of the determination result data 202 to display a defect determination result. Various types of display formats are usable. For example, only the presence or absence of a defect in an inspection target area can be displayed, and the position of the defect can be displayed based on marker information. In addition, the size of the defect can be displayed based on the area (the number of pixels) of a dark part.
[0325] As described above, in the present embodiment, similarly to the first embodiment, a defect which needs to be detected can be reliably detected without misrecognizing any irregularity of a normal ceramic surface as a defect by using the minimum luminance image data generated from a plurality of pieces of captured image data obtained in different irradiation directions of the illumination light.
[0326] In particular, in the defect inspection apparatus according to the present embodiment, the small-angle lighting part having an illumination light irradiation angle which is smaller than the irradiation angle of the first lighting part, in addition to the medium-angle lighting part and the large-angle lighting part, which are the first illumination part and the second illumination part in the defect inspection apparatus according to the first embodiment, and the determination processing based on the small-angle captured image data obtained by the image pickup using the small-angle illumination part is provided in addition to the determination processings using the medium-angle -illumination part and the large-angle illumination part, which correspond to the first determination processing and the second determination processing in the first embodiment, are performed, whereby the examination in ei is accurately performed in an inspection target area larger than that of the defect inspection apparatus according to the first embodiment.
[0327] Thus, the defect inspecting device according to the present embodiment can perform preferential inspection in an inspection target area having an area several tens of times that of the defect inspecting device according to the first embodiment, in some cases in the comparison where the distance from the camera to the front side of the honeycomb structural body is substantially constant. In such a case, the inspection for an extremely large inspection target area can be accurately performed without increasing the size of the defect inspection apparatus. For example, when the inspection target area of the defect inspection apparatus according to the first embodiment is a square area of 10mm x 10mm = 100mm 2 and the inspection target area of the defect inspection apparatus according to the present embodiment is a square area of 60 mm x 60 mm = 3600 mm 2 is, the latter inspection target area is 36 times as large as the former inspection target area. Modification of the second embodiment
[0328] In the second embodiment described above, image pickup is performed using the small-angle illumination part 115 , an image capture using the medium-angle lighting part 120 and image capturing using the wide-angle lighting part 130 in the order given as in figure 32, but this order can be changed. In this case, the in figure 36 of the corresponding determination processing can be changed.
[0329] With a relatively small size of the honeycomb structural body 1 can capture the image using the small-angle lighting part 115 and the series of processings up to the following small-angle determination processing based on the small-angle captured image data obtained by the image capture in the flaw inspection with the flaw inspection device 3000 may be omitted as long as the accuracy of the study is maintained. In other words, the defect inspection device 3000 can as the defect inspection device 1000 can be used according to the first embodiment.
[0330] Similar to the inspecting method according to the first embodiment, the inspecting method according to the second embodiment is also for detecting as a convex part (for example, a protrusion) on the face 1a of the honeycomb structural body 1 existing defect usable.
[0331] Specifically, in the second embodiment described above, the determination image generation part generates 230Small-angle (Middle-angle, Big-angle) minimum luminance image data in which a minimum value of luminance values of a plurality of corrected small-angle (Middle-angle, Big-angle) image data at an identical pixel position as a luminance value is set at the pixel position, and generates small-angle (middle-angle, big-angle) detection image data on the basis of the small-angle (middle-angle, big-angle) minimum luminance image data, but instead, it may be small-angle (middle-angle, big-angle) maximum luminance image data in which a maximum value of luminance values of the plurality of corrected small-angle (middle-angle, big-angle) image data is set at an identical pixel position as a luminance value at the pixel position, and generate the small-angle (mean-angle, big-angle) detection image data on the basis of the small-angle (mean -angle, large-angle-) generate maximum luminance image data. In this case, the small-angle (middle-angle, big-angle) sets maximum luminance image data by practically overlaying areas with high luminance values in the pieces of small-angle (middle-angle, big-angle) Thus, the small-angle (middle-angle, large-angle) maximum luminance image data is data in which an area having a high luminance value due to a convex defect is enhanced.
[0332] Then, when an area of bright pixels in an area greater than or equal to a predetermined bright-part threshold in one of the small-angle detection image data and the medium-angle detection image data as well as the large -angle determination image data is present, the defect determination part determines 240 that there is a convex defect at the occurrence position of the bright pixel area. Accordingly, a convex defect present on the outer surface of a ceramic body, which needs to be detected, can be reliably detected without misrecognizing an irregularity of a normal ceramic surface as a defect.
[0333] In the second embodiment described above, each dimming unit is individually dimmed to each of the small-angle lighting elements 116 , the mid-angle lighting elements 121 and the high-angle lighting elements 131 possible, but with the high-angle lighting elements 131 and also in the mid-angle lighting elements 121 the individual dimming function can be dispensed with. The reason is that the distance dependence of the luminance in the small-angle lighting elements 116 is most evident and a substantially constant luminance is likely to be achieved without individual dimming in the viewing angle as the irradiance angle increases as with the mid-angle illuminants 121 and the high-angle lighting elements 131 .
[0334] In the second embodiment described above, each of the small-angle lighting elements includes 116 , the mid-angle lighting elements 121 and the high-angle lighting elements 131 two dimming units, but instead each may contain a larger number of dimming units, each individually dimmable. Other changes
[0335] In the above-described embodiments, the inspection target is the face 1a of the honeycomb structural body 1 , but in an aspect similar to the aspect described above, the inspection is by the defect inspection devices 1000 and 2000 also applicable to a macroscopically horizontal ceramic surface (surface of a ceramic body) having minute irregularities.
[0336] In the above-described embodiments, a honeycomb structural body including a quadrangular prism-shaped honeycomb (square in cross section) is exemplified as an inspection target, but a honeycomb structural body including a hexagonal-columnar honeycomb (hexagonal in cross section) may also be an inspection target, or honeycombs having a pentagonal columnar shape, honeycomb structural bodies including an octagonal columnar shape and other various kinds of shapes can be objects of study. QUOTES INCLUDED IN DESCRIPTION
[0000] This list of documents cited by the applicant was generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions. Patent Literature Cited
[0000] JP 2010249798
[0009] JP 2008139052
[0009]
Claims
[1] Ceramic body defect detection device configured to detect the presence of a defect on an outer surface of a ceramic body, the device comprising: a table on which a ceramic object is to be placed as the object of investigation; an image acquisition unit configured to capture an image of a target area as at least a part of a target surface of the ceramic body placed on the table in a normal direction to the target area; one or more lighting units, comprising four or more lighting elements configured to illuminate the target area of the investigation with light at an identical irradiation angle in different irradiation directions arranged at equal angular intervals around the image acquisition unit; an investigation image generation unit configured to generate investigation image data for determining the presence of a defect in the investigation target area, based on image data acquired by the image acquisition unit; and a defect detection component configured to determine the presence of a defect in the target area of investigation based on the detection image data, wherein the numerous lighting elements are switched on and off one after the other, The image acquisition part, by capturing an image of the target area when each of the many lighting elements is switched on, generates a multitude of pieces of image data and the investigative image generation component: Minimum luminance image data is generated in which a minimum value of luminance values of the multitude of pieces of acquisition image data at an identical pixel position is set as a luminance value at the pixel position, and then the determination image data is generated based on the minimum luminance image data; or Maximum luminance image data is generated, in which a maximum value of luminance values of the multitude of pieces of recording image data at an identical pixel position is set as a luminance value at the pixel position, and then the determination image data is generated based on the maximum luminance image data. [2] Ceramic body defect detection device according to claim 1, wherein the image generation part furthermore, it includes a luminance correction processing section configured to correct the luminance values of the multitude of pieces of image data so that the luminance values at a reference piece previously defined in each of the multitude of image data pieces are the same for the multitude of image data pieces, and The minimum luminance image data or the maximum luminance image data is generated based on the multitude of pieces of captured image data whose luminance values are corrected by the luminance correction processing part. [3] Ceramic body defect detection device according to claim 1 or 2, wherein one or more illumination parts include: a first illumination section, which comprises a multitude of first illumination elements, each configured to illuminate the target area with light at a first irradiation angle of 30° to 60° inclusive; and a second lighting component, which, as a multitude of lighting elements, contains a multitude of second lighting elements, each configured to irradiate the investigation target area with illumination light at a second irradiation angle of 60° to 85° inclusive. [4] Ceramic body defect detection device according to claim 3, wherein The image acquisition unit generates a multitude of pieces of first image data by taking an image of the target area for each of the multitude of first illumination elements when each of the multitude of first illumination elements is switched on, and generates a multitude of pieces of second image data by taking an image of the target area for each of the multitude of second illumination elements when each of the multitude of second illumination elements is switched on. the multitude of pieces of first image data and the multitude of pieces of second image data are the multitude of pieces of different image data. In the event that the investigation image data is to be generated based on the minimum luminance image data, the investigation image generation part consists of the first minimum luminance image data, in which a minimum value of luminance values of the multitude of pieces of first image data at an identical pixel position is set as a luminance value at the pixel position, and the second minimum luminance image data, in which a minimum value of luminance values of the multitude of pieces of second image data at an identical pixel position is set as a luminance value at the pixel position, as the minimum luminance image data is generated and the investigation image data is generated by generating first investigation image data based on the first minimum luminance image data and generating second investigation image data based on the second minimum luminance image data; and In the event that the investigation image data is to be generated based on the maximum luminance image data, the investigation image generation part consists of the first maximum luminance image data, in which a maximum value of luminance values of the multitude of pieces of the first image data at an identical pixel position is set as a luminance value at the pixel position, and the second maximum luminance image data, in which a maximum value of luminance values of the multitude of pieces of the second image data at an identical pixel position is set as a luminance value at the pixel position, as the maximum luminance image data is generated and the investigation image data is generated by generating the first investigation image data based on the first maximum luminance image data and generating the second investigation image data based on the second maximum luminance image data. The defect detection part determines the presence of a defect in the target area of investigation based on both the first and second investigation image data. [5] Ceramic body defect detection device according to claim 3, wherein the number of first lighting elements is equal to the number of second lighting elements, and the first and second lighting elements are arranged such that a vertical plane containing an illumination direction of each of the first lighting elements always contains an illumination direction of one of the second lighting elements. the multitude of first lighting elements emits illumination light belonging to a first wavelength band, and the multitude of second lighting elements emits illumination light belonging to a second wavelength band different from the first wavelength band. Each pair of the first and second lighting elements with irradiation directions contained in an identical vertical plane is switched on and off simultaneously, The image acquisition section generates the multitude of pieces of image data by capturing an image of the target area when each pair of the first and second illumination elements is switched on. The defect detection device further includes a separation image generation part which is configured to generate a plurality of pieces of first separation image data and a plurality of pieces of second separation image data by performing a color separation of each from the plurality of pieces of acquisition image data based on the first wavelength band and the second wavelength band. In the event that the investigation image data is to be generated based on the minimum luminance image data, the investigation image generation part consists of the first minimum luminance image data, in which a minimum value of luminance values of the multitude of pieces of first separation image data at an identical pixel position is set as a luminance value at the pixel position, and the second minimum luminance image data, in which a minimum value of luminance values of the multitude of pieces of second separation image data at an identical pixel position is set as a luminance value at the pixel position, as the minimum luminance image data is generated and the investigation image data is generated by generating first investigation image data based on the first minimum luminance image data and generating second investigation image data based on the second minimum luminance image data; and In the case that the investigation image data is to be generated based on the maximum luminance image data, the investigation image generation part consists of the first maximum luminance image data, in which a maximum value of luminance values of the multitude of pieces of first separation image data at an identical pixel position is set as a luminance value at the pixel position, and the second maximum luminance image data, in which a maximum value of luminance values of the multitude of pieces of second separation image data at an identical pixel position is set as a luminance value at the pixel position, when the maximum luminance image data is generated and the investigation image data is generated by generating the first investigation image data based on the first maximum luminance image data and generating the second investigation image data based on the second maximum luminance image data. The defect detection part determines the presence of a defect in the target area of investigation based on both the first and second investigation image data. [6] Ceramic body defect detection device according to claim 1 or 2, wherein one or more illumination parts include: a small-angle illumination component, which, as the multitude of illumination elements, contains a multitude of small-angle illumination elements, each configured to irradiate the investigation target area with illumination light at an irradiation angle of 5° to 30° inclusive; a mid-angle illumination element, which, as a plurality of illumination elements, contains a plurality of mid-angle illumination elements, each configured to illuminate the study target area with light at an irradiation angle of 30° to 60° inclusive; and a wide-angle lighting element, which, as the multitude of lighting elements, contains a multitude of wide-angle lighting elements, each configured to irradiate the target area of investigation with illumination light at an irradiation angle of 60° to 85° inclusive. [7] Ceramic body defect detection device according to claim 6, wherein The image acquisition unit generates a multitude of small-angle image data pieces by taking an image of the target area for each of the multitude of small-angle illumination elements when each of the multitude of small-angle illumination elements is switched on; generates a multitude of medium-angle image data pieces by taking an image of the target area for each of the multitude of medium-angle illumination elements when each of the multitude of medium-angle illumination elements is switched on; and generates a multitude of large-angle image data pieces by taking an image of the target area for each of the multitude of large-angle illumination elements when each of the multitude of large-angle illumination elements is switched on. the multitude of small-angle image data pieces, the multitude of medium-angle image data pieces, the multitude of large-angle image data pieces, the multitude of different image data pieces are, In the case that the investigation image data is to be generated based on the minimum luminance image data, the investigation image generation part consists of small-angle minimum luminance image data in which a minimum value of luminance values of the multitude of pieces of small-angle image data at an identical pixel position is set as a luminance value at the pixel position, medium-angle minimum luminance image data in which a minimum value of luminance values of the multitude of pieces of medium-angle image data at an identical pixel position is set as a luminance value at the pixel position, and large-angle minimum luminance image data in which a minimum value of luminance values of the multitude of pieces of large-angle image data at an identical pixel position is set as a luminance value at the pixel position.when the minimum luminance image data is generated and the detection image data is generated by generating small-angle detection image data based on the small-angle minimum luminance image data, generating medium-angle detection image data based on the medium-angle minimum luminance image data, and generating large-angle detection image data based on the large-angle minimum luminance image data; and , In the case that the investigation image data is to be generated based on the maximum luminance image data, the investigation image generation part is small-angle maximum luminance image data, in which a maximum value of luminance values of the multitude of pieces of small-angle image data at an identical pixel position is set as a luminance value at the pixel position; medium-angle maximum luminance image data, in which a maximum value of luminance values of the multitude of pieces of medium-angle image data at an identical pixel position is set as a luminance value at the pixel position; and large-angle maximum luminance image data, in which a maximum value of luminance values of the multitude of pieces of large-angle image data at an identical pixel position is set as a luminance value at the pixel position.when the maximum luminance image data is generated and the determination image data is generated by generating small-angle determination image data based on the small-angle maximum luminance image data, generating medium-angle determination image data based on the medium-angle maximum luminance image data, and generating large-angle determination image data based on the large-angle maximum luminance image data. The defect detection part determines the presence of a defect in the target area of investigation based on all of the small-angle detection image data, the medium-angle detection image data and the large-angle detection image data. [8] Ceramic body defect detection device according to any one of claims 1 to 7, wherein the investigation image generation part generates the investigation image data as binary data and In the event that the investigation image data is generated based on the minimum luminance image data, the defect detection part determines that a defect is present in the investigation target area if the investigation image data contains a dark area on a surface that is greater than or equal to a predetermined first dark area threshold, and In the case that the detection image data is generated based on the maximum luminance image data, the defect detection part determines that a defect is present in the target area if the detection image data contains a bright area on a surface which is greater than or equal to a predetermined bright area threshold. [9] Ceramic body defect detection device according to claim 8, wherein In the case that the ceramic body is a sealed honeycomb structure and the target surface of the investigation is an end face of the sealed honeycomb structure, the investigation image generation part generates the investigation image data based on the minimum luminance image data and The defect detection component also determines that a defect is present in the target area of investigation if, in the detection image data in which a dark part corresponding to an opening in the target area is excluded and a light part corresponding to any connecting part and a dark part corresponding to the outside of the ceramic body, if present, are additionally excluded, a dark part is present on an area which is greater than or equal to a predetermined second dark part threshold. [10] Ceramic body defect detection device according to any one of claims 3 to 5, wherein the multitude of primary lighting elements and the multitude of secondary lighting elements are held by a single mounting body and the multitude of first lighting elements is arranged in one plane and the multitude of second lighting elements is arranged in another plane. [11] Ceramic body defect detection device according to claim 6 or 7, wherein the plurality of small-angle lighting elements each contains at least two individually dimmable dimming units. [12] Ceramic body defect detection device according to claim 11, wherein the multitude of small-angle lighting elements, the multitude of medium-angle lighting elements, and the multitude of large-angle lighting elements are held by a single mounting body and the multitude of small-angle lighting elements, the multitude of medium-angle lighting elements, and the multitude of large-angle lighting elements are arranged in their respective different planes. [13] Ceramic body defect detection device according to one of claims 1 to 12, wherein the plurality of lighting elements contained in each of the one or more lighting parts consists of eight lighting elements. [14] Method for examining for the presence of a defect on an outer surface of a ceramic body, the method comprising: a placement step of placing a ceramic body as the object of investigation on a predetermined table; an image acquisition step of generating a multitude of pieces of image data by capturing, using a predetermined image acquisition device, an image of a target area as at least a part of a target surface of the ceramic body placed on the table in a normal direction of the target area; an investigative image generation step of generating investigative image data to determine the presence of a defect in the target area of investigation based on the multitude of pieces of image data; and a defect detection step of determining the presence of a defect in the target area of investigation based on the investigation image data, wherein In the image acquisition step, the multitude of pieces of image data is generated by successively switching on and off four or more illumination elements, which are provided in one or more illumination parts and configured to emit illumination light obliquely at an identical irradiation angle in respective irradiation directions arranged at equal angular intervals around the image acquisition device, and by capturing an image of the target area when each of the multitude of illumination elements is switched on. in the investigative image generation step Minimum luminance image data, in which a minimum value of luminance values of the multitude of pieces of image data at an identical pixel position as a luminance value at the pixel position is generated, and then the determination image data is generated based on the minimum luminance image data; or Maximum luminance image data, in which a maximum value of luminance values of the multitude of pieces of recording image data at an identical pixel position is set as a luminance value at the pixel position, are generated, and then the determination image data are generated on the basis of the maximum luminance image data. [15] Ceramic body defect detection method according to claim 14, wherein the detection image generation step further comprises a luminance correction processing step of correcting luminance values of the plurality of pieces of image data such that the luminance values at a reference part previously defined in each of the plurality of image data pieces are the same in the plurality of image data pieces; and In the image generation step, the minimum luminance image data or the maximum luminance image data are generated based on the multitude of pieces of image data whose luminance values are corrected by the luminance correction processing step. [16] Ceramic body defect detection method according to claim 14 or 15, wherein one or more illumination parts include: a first illumination section, which comprises a multitude of first illumination elements, each configured to illuminate the target area with light at a first irradiation angle of 30° to 60° inclusive; and a second lighting component, which, as a multitude of lighting elements, contains a multitude of second lighting elements, each configured to irradiate the investigation target area with illumination light at a second irradiation angle of 60° to 85° inclusive. [17] Ceramic body defect detection method according to claim 16, wherein In the image acquisition step, a multitude of pieces of first acquisition image data are generated by taking an image of the target area for each of the multitude of first illumination elements when each of the multitude of first illumination elements is switched on, and a multitude of pieces of second acquisition image data are generated by taking an image of the target area for each of the multitude of second illumination elements when each of the multitude of second illumination elements is switched on. the multitude of pieces of first image data and the multitude of pieces of second image data are the multitude of pieces of different image data. in the investigative image generation step In the case of generating the investigation image data based on the minimum luminance image data, the first minimum luminance image data, in which a minimum value of luminance values of the multitude of pieces of the first image data at an identical pixel position is set as a luminance value at the pixel position, and the second minimum luminance image data, in which a minimum value of luminance values of the multitude of pieces of the second image data at an identical pixel position is set as a luminance value at the pixel position, are generated when the minimum luminance image data are generated and the investigation image data are generated by generating the first investigation image data based on the first minimum luminance image data and generating the second investigation image data based on the second minimum luminance image data; and In the case of generating the investigation image data based on the maximum luminance image data, the first maximum luminance image data, in which a maximum value of luminance values of the multitude of pieces of the first image data at an identical pixel position is set as a luminance value at the pixel position, and the second maximum luminance image data, in which a maximum value of luminance values of the multitude of pieces of the second image data at an identical pixel position is set as a luminance value at the pixel position, are generated when the maximum luminance image data are generated and the investigation image data are generated by generating the first investigation image data based on the first maximum luminance image data and generating the second investigation image data based on the second maximum luminance image data. In the defect detection step, the presence of a defect in the target area of investigation is determined based on both the first and second investigation image data. [18] Ceramic body defect detection method according to claim 16, wherein the number of first lighting elements is equal to the number of second lighting elements, and the first and second lighting elements are arranged such that a vertical plane containing an irradiation direction of each of the first lighting elements always contains an irradiation direction of one of the second lighting elements, and then the multitude of first lighting elements emits illuminating light belonging to a first wavelength band, and the multitude of second lighting elements emits illuminating light belonging to a second wavelength band different from the first wavelength band. Each pair of the first and second lighting elements with irradiation directions contained in an identical vertical plane is switched on and off simultaneously, In the image acquisition step, the multitude of pieces of image data is generated by capturing an image of the target area when each pair of the first and second illumination elements is switched on. The defect detection procedure further includes a separation image generation step of generating a plurality of pieces of first separation image data and a plurality of pieces of second separation image data by performing a color separation of each of the plurality of pieces of acquisition image data based on the first wavelength band and the second wavelength band. in the investigative image generation step In the case of generating the investigation image data based on the minimum luminance image data, the first minimum luminance image data, in which a minimum value of luminance values of the multitude of pieces of first separation image data at an identical pixel position is set as a luminance value at the pixel position, and the second minimum luminance image data, in which a minimum value of luminance values of the multitude of pieces of second separation image data at an identical pixel position is set as a luminance value at the pixel position, are generated when the minimum luminance image data are generated and the investigation image data are generated by generating the first investigation image data based on the first minimum luminance image data and generating the second investigation image data based on the second minimum luminance image data; and In the case of generating the investigation image data based on the maximum luminance image data, the first maximum luminance image data, in which a maximum value of luminance values of the multitude of pieces of the first separation image data at an identical pixel position is set as a luminance value at the pixel position, and the second maximum luminance image data, in which a maximum value of luminance values of the multitude of pieces of the second separation image data at an identical pixel position is set as a luminance value at the pixel position, are generated when the maximum luminance image data are generated and the investigation image data are generated by generating the first investigation image data based on the first maximum luminance image data and generating the second investigation image data based on the second maximum luminance image data. In the defect detection step, the presence of a defect in the target area of investigation is determined based on both the first and second investigation image data. [19] Ceramic body defect detection method according to claim 14 or 15, wherein one or more illumination parts include: a small-angle illumination component, which, as the multitude of illumination elements, contains a multitude of small-angle illumination elements, each configured to irradiate the investigation target area with illumination light at an irradiation angle of 5° to 30° inclusive; a mid-angle illumination element, which, as a plurality of illumination elements, contains a plurality of mid-angle illumination elements, each configured to illuminate the study target area with light at an irradiation angle of 30° to 60° inclusive; and a wide-angle lighting element, which, as the multitude of lighting elements, contains a multitude of wide-angle lighting elements, each configured to irradiate the target area of investigation with illumination light at an irradiation angle of 60° to 85° inclusive. [20] Ceramic body defect detection method according to claim 19, wherein In the image acquisition step, a multitude of small-angle image data pieces are generated by taking an image of the target area for each of the multitude of small-angle illumination elements when each of the multitude of small-angle illumination elements is switched on; a multitude of medium-angle image data pieces are generated by taking an image of the target area for each of the multitude of medium-angle illumination elements when each of the multitude of medium-angle illumination elements is switched on; and a multitude of large-angle image data pieces are generated by taking an image of the target area for each of the multitude of large-angle illumination elements when each of the multitude of large-angle illumination elements is switched on. the multitude of small-angle image data pieces, the multitude of medium-angle image data pieces, the multitude of large-angle image data pieces, the multitude of different image data pieces are, in the investigative image generation step; In the case of generating the investigation image data based on the minimum luminance image data, small-angle minimum luminance image data in which a minimum value of luminance values of the multitude of pieces of small-angle image data at an identical pixel position is set as a luminance value at the pixel position, medium-angle minimum luminance image data in which a minimum value of luminance values of the multitude of pieces of medium-angle image data at an identical pixel position is set as a luminance value at the pixel position, and large-angle minimum luminance image data in which a minimum value of luminance values of the multitude of pieces of large-angle image data at an identical pixel position is set as a luminance value at the pixel position, when the minimum luminance image data are generated,the investigation image data are generated by creating small-angle investigation image data based on small-angle minimum luminance image data, creating medium-angle investigation image data based on medium-angle minimum luminance image data, and creating large-angle investigation image data based on large-angle minimum luminance image data; and , In the case of generating the investigation image data based on the maximum luminance image data, small-angle maximum luminance image data in which a maximum value of luminance values of the multitude of pieces of small-angle image data at an identical pixel position is set as a luminance value at the pixel position, medium-angle maximum luminance image data in which a maximum value of luminance values of the multitude of pieces of medium-angle image data at an identical pixel position is set as a luminance value at the pixel position, and large-angle maximum luminance image data in which a maximum value of luminance values of the multitude of pieces of large-angle image data at an identical pixel position is set as a luminance value at the pixel position,when the maximum luminance image data are generated and the determination image data are generated by generating small-angle determination image data based on the small-angle maximum luminance image data, generating medium-angle determination image data based on the medium-angle maximum luminance image data and generating large-angle determination image data based on the large-angle maximum luminance image data and , In the defect detection step, the presence of a defect in the target area is determined based on all of the small-angle detection image data, the medium-angle detection image data and the large-angle detection image data. [21] Ceramic body defect detection method according to one of claims 14 to 20, wherein In the investigation image generation step, the investigation image data is generated as binary data and in the defect detection step In the event that the investigation image data are generated based on the minimum luminance image data, it is determined that a defect exists in the investigation target area if the investigation image data contains a dark area on a surface that is greater than or equal to a predetermined first dark area threshold; and In the case where the investigation image data is generated based on the maximum luminance image data, it is determined that a defect exists in the target area of investigation if the investigation image data contains a bright area on a surface which is greater than or equal to a predetermined bright area threshold. [22] Ceramic body defect detection method according to claim 21, wherein In the case that the ceramic body is a sealed honeycomb structure and the target surface of the investigation is an end face of the sealed honeycomb structure, In the investigation image generation step, the investigation image data is generated based on the minimum luminance image data and In the defect detection step, even if the detection image data, in which a dark part corresponding to an opening in the target area is excluded and a light part corresponding to any connecting part and a dark part corresponding to the outside of the ceramic body, if present, are additionally excluded, a dark part is present on an area which is greater than or equal to a predetermined second dark part threshold, it is determined that a defect is present in the target area. [23] Ceramic body defect detection method according to claim 19 or 20, wherein the multitude of small-angle lighting elements each contains at least two individually dimmable dimming units and A luminance difference corresponding to the difference between distances from the multitude of small-angle lighting elements in a recording area of the predetermined image recording device is reduced by individually dimming the at least two dimming units in advance of the image recording step by the predetermined image recording device. [24] Ceramic body defect detection method according to one of claims 14 to 23, wherein the plurality in each of the one or more lighting elements provided consists of eight lighting elements.