JP2007333579A - Inspection method and system for honeycomb structure - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To more stably perform high-accuracy inspection in inspecting clogging of honeycomb structure using an image processing method. <P>SOLUTION: In this inspection method, in order to inspect cell clogging in the honeycomb structure 1, illuminating light is transmitted through the cell 11 by an illuminating device 2, in the state of illuminating the honeycomb structure from one end side, the end face on the other side of the honeycomb structure is photographed by a camera 3 to obtain an inspection image, and whether cell clogging is good or bad is determined based on the inspection image. The number of through cells where the illuminating light is transmitted without interception or the illuminating light is partially transmitted is measured from the inspection image, and the quality is determined based on the number of through cells. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to an inspection technique for a honeycomb structure, and more particularly to an inspection technique for inspecting clogging of cells using an end face image of the honeycomb structure.

  A honeycomb structure is a structure in which a large number of through-pore cells having a cross-sectional shape of a hexagon, a quadrangle, or a circle are formed in a honeycomb shape, and is used as a carrier for a catalyst for exhaust gas purification. When such a honeycomb structure is used as a catalyst carrier, if the cells are clogged, the catalyst capacity is lowered at that portion. Therefore, in the honeycomb structure, cell clogging is a particular problem, and a technique for efficiently inspecting cell clogging is required. Here, in the case of clogging of cells, clogging includes clogging that occurs in the manufacturing process of the honeycomb structure itself and clogging that occurs in the process of supporting the catalyst on the honeycomb structure.

  As for clogging inspection of a honeycomb structure, a visual method and a method using image processing are known. In the visual method, a person visually checks the end face image of the honeycomb structure to inspect clogging (for example, Patent Document 1). There are two methods for obtaining the end face image. In one method, an end face image of the honeycomb structure is formed on an imaging plate arranged on the other end face side by illumination light irradiated so as to pass through the cells from one end face side of the honeycomb structure. In another method, an end face image is obtained as an end face image obtained by photographing the other end face in a state where illumination light is irradiated so as to pass through the cells from one end face side of the honeycomb structure.

  On the other hand, in the image processing method, an end face image acquired as described above is used as an inspection image, and clogging is determined from the inspection image by image processing. Specifically, utilizing the fact that clogged cells darken in the inspection image, the end face image of the honeycomb structure in the inspection image is divided into a bright part and a dark part, and the area ratio of the dark part to the entire end face image Ask for. And the quality regarding clogging is determined based on the area ratio (for example, patent document 2).

  In addition to such a clogging inspection, a technique for inspecting a honeycomb structure by an image processing method is known. For example, a technique (for example, Patent Document 3) for inspecting the unevenness of the partition wall surface of a cell using an end face image as an inspection image, a technique for inspecting an end face of a honeycomb structure using an end face image as an inspection image (for example, Patent Document 4), or an end face image Is a technique (for example, Patent Document 5) for inspecting the outer shape of the honeycomb structure.

JP-A-6-43114 Japanese Patent Application Laid-Open No. 5-26459 JP 2005-134140 A JP 2002-257736 A JP 2002-267427 A

  The image processing method described above makes it possible to automate clogging inspection. However, the conventional image processing method still has a problem in the reliability of the inspection, and for this reason, the automation of the inspection by the image processing method has not actually progressed, and the actual condition of the clogging inspection is still based on the visual method. It is.

  The reason why the conventional image processing method leaves a problem in the inspection reliability is due to the use of the area ratio. As described above, the conventional image processing method relates to clogging by the area ratio of clogged cells obtained by dividing the end face image in the inspection image into the bright part and the dark part using the fact that the clogged cells become dark in the inspection image. The quality is judged. Therefore, it is affected by the partition walls of the cell. Honeycomb structures are mainly made of ceramics and metal, and particularly in the case of metal honeycomb structures, there are many cases in which the thickness of the partition walls of the cells varies. The partition walls of the cells appear in the inspection image as dark portions like the clogged cells. Therefore, if the thickness of the partition walls of the cell varies, the dark area changes regardless of the clogged cell, and the accuracy of the inspection becomes unstable, causing a problem in the reliability of the inspection.

  The present invention has been made against the background described above, and it is an object of the present invention to enable more stable and high-precision inspection for clogging inspection of a honeycomb structure by an image processing method.

  In the present invention, in order to solve the above problems, in order to inspect the clogging of the cells in the honeycomb structure in which a large number of cells are formed in a honeycomb shape, illumination light is transmitted through the cells by an illumination device. In the state where the honeycomb structure is illuminated from one end side, the end face on the other end side of the honeycomb structure is photographed to obtain an inspection image, and quality determination regarding the cell clogging is performed based on the inspection image. In the inspection method described above, the number of penetrating cells, which are the cells through which the illumination light from the illumination device passes without being blocked or the illumination light from the illumination device partially transmits, is the inspection image. And the pass / fail judgment is made based on the number of penetrating cells.

  Thus, by measuring the number of penetrating cells from the inspection image and making a pass / fail determination regarding clogging based on the number of penetrating cells, the influence of the cell partition wall image on the inspection image can be effectively avoided. Therefore, the clogging inspection of the honeycomb structure by the image processing method can be stably performed with high accuracy.

  In the present invention, in the inspection method as described above, the penetration cell rate is obtained by dividing the number of penetration cells by the total number of cells in the honeycomb structure, and the pass / fail judgment is performed based on the penetration cell rate. ing. By doing in this way, the quality determination based on the number of penetration cells can be performed more effectively.

  According to the present invention, in the inspection method as described above, the number of through-cells is measured based on a partition removal image generated by removing an image of the partition walls of the cells from the inspection image. By measuring the number of penetrating cells based on the barrier rib removal image in this way, it is possible to avoid the barrier rib image from affecting the measurement of the number of penetrating cells when measuring the number of penetrating cells. Measurement accuracy can be increased.

  Further, in the present invention, for the inspection method as described above, the number of completely penetrating cells, which are cells that are transmitted without being blocked by the illumination light from the illumination device, is measured from the inspection image, and the number of completely penetrating cells is calculated as described above. By dividing by the total number of cells in the honeycomb structure, the complete penetration cell rate is obtained, and the complete penetration cell rate can be used for the quality determination.

  Further, according to the present invention, in the inspection method as described above, the through cell configuration ratio is obtained by dividing the number of complete through cells by the number of through cells, and the through cell configuration ratio can be used for the pass / fail judgment. ing.

  Further, in the present invention, for the inspection method as described above, the number of partially penetrating cells that are cells through which the illumination light from the illumination device partially transmits is measured from the inspection image, and the number of the partially penetrating cells is determined as the honeycomb structure. The partial penetration cell rate is obtained by dividing by the total number of cells in the body, and the partial penetration cell rate can be used for the quality determination.

  As described above, by making it possible to use the complete penetration cell rate, the penetration cell composition ratio, or the partial penetration cell rate for the pass / fail judgment, it is possible to diversify the inspection contents and to enable a more advanced inspection. .

  Further, in the present invention, for the inspection method as described above, the end face of the honeycomb structure is divided into a plurality of divided inspection regions, the inspection images are obtained for each of the divided inspection regions, and the inspection images are obtained based on the plurality of inspection images. Pass / fail judgment can be made. By doing so, it is possible to prevent the cell image in the inspection image from becoming too small, and to perform the clogging inspection more effectively.

  According to the present invention as described above, the clogging inspection of the honeycomb structure by the image processing method can be performed more stably and with high accuracy.

  Hereinafter, modes for carrying out the present invention will be described. FIG. 1 schematically shows a hardware configuration of a honeycomb structure inspection system according to an embodiment. The honeycomb structure inspection system S of the present embodiment provides illumination light in a state in which each of a large number of cells 11 in the subject 1 is transmitted from one end surface (illumination side end surface) of the subject (honeycomb structure) 1. The illumination device 2 that irradiates the object 1, the camera 3 that images the subject 1 illuminated by the illumination device 2 with respect to the other end surface (end surface on the imaging side), and the position of the camera 3 in relation to the imaging side end surface The cell penetration rate as described later is obtained from the end face image of the subject 1 obtained by the XY stage 4 and the camera 3, ie, the inspection image, which is obtained by the camera position control means, and based on the cell penetration rate, The determination processing device 5 that performs pass / fail determination, the display device 6 that displays the determination result by the determination processing device 5, and various information of the subject 1 that is necessary when examining the subject 1 are provided to the determination processing device 5. And a server 7 which is the subject information storage unit.

  Among these inspection system components, the XY stage 4 and the server 7 are not necessarily indispensable functional elements for the minimum clogging inspection of the subject 1, so that the clogging inspection can be performed more effectively. Is a functional element. That is, the XY stage 4 enables the clogging inspection more effectively by enabling the divided photographing inspection. Here, the division imaging inspection is a technique in which the imaging side end face of the subject 1 is divided into a plurality of divided inspection areas, and an inspection image is imaged for each of the divided inspection areas. When the end face size of the subject 1 is large, if the entire end face of the subject 1 is taken into one inspection image, the image of the cell 11 in the inspection image becomes too small, and the clogging inspection due to the cell penetration rate is effective. May not be possible. In such a case, a division photographing inspection is used. In the division imaging examination, it is necessary to perform movement / position control of the camera 3 in accordance with the division examination area set and divided on the imaging side end face of the subject 1. In addition, in the divided imaging inspection, it is important to accurately grasp in which divided inspection area the cell located at the boundary of the divided inspection area is measured. For this purpose, the movement / position control of the camera 3 with respect to each divided inspection area. High accuracy is required. The XY stage 4 makes it possible to perform divided photographing inspections by responding to such high-precision movement / position control requests of the camera 3.

  On the other hand, the server 7 provides various pieces of subject information related to the clogging inspection of the subject 1, so that the clogging inspection can be more effectively performed even when the subject 1 is a honeycomb structure of a different type. Like that. Here, the subject information related to the clogging inspection includes subject attribute information such as the width, height, number of constituent cells, and cell shape of the subject 1, and inspection item information such as the inspection range and pass / fail judgment value. In addition, there are inspection condition parameter information such as coefficients described later. The server 7 functions to store such object information in advance for the honeycomb structure that may be a target in the inspection system, and to provide it to the determination processing device 5 when inspecting the object 1. The function enables the clogging inspection to be performed more effectively by making it possible to easily cope with a change in the type of the subject 1.

  FIG. 2 shows a configuration example of the determination processing device 5. The determination processing device 5 of this example includes an inspection processing control unit 51 that controls overall inspection processing such as determination of pass / fail of the subject 1 based on the cell penetration rate, and determines the pass / fail of the subject 1 by obtaining the cell penetration rate from the inspection image. An inspection processing unit 52 that performs an inspection process such as determining an inspection, an external input / output unit 53 that controls input / output with the camera 3, the display device 6, and the like A system bus 54 that transmits data and a processing data storage unit 55 that stores various types of data associated with processing in the inspection processing unit 52 are provided. The processing data storage unit 55 includes an object information storage unit 56 that stores object information, an inspection result storage unit 57 that stores a processing result in the inspection processing unit 52, and an inspection image storage unit 58 that stores an inspection image. It has.

  The above is the system configuration of the honeycomb structure inspection system S. Below, the inspection method performed with the honeycomb structure inspection system S is demonstrated. FIG. 3 shows a processing flow in the inspection processing by the honeycomb structure inspection system S. The inspection processing includes subject information capturing processing 101 for capturing subject information from the server 7, inspection image capturing processing 102 for capturing and capturing an inspection image with the camera 3, and camera movement determination processing 103 for determining whether or not to move the camera. Based on the partition extraction image and the inspection image obtained by the camera movement processing 104 for outputting the control signal to the XY stage 4 to move the camera 3, the partition extraction processing 105 for extracting the cell partition from the inspection image, and the partition extraction processing 105. Penetrating cell extraction processing 106 for extracting penetrating cells, and extraction end determination processing for determining whether penetrating cell extraction has been completed for all of a plurality of divided examination regions set on the end face of the subject 1 when divided imaging examination is performed. 107, the cell penetration rate measurement processing 108 for obtaining the cell penetration rate based on the processing result in the penetration cell extraction processing 106, and the subject 1 from the cell penetration rate It includes acceptability determination process 109 for performing determination. Details of each of these processes will be described below.

  In the subject information acquisition process 101, the subject information is acquired from the server 7 and stored in the subject information storage unit 56. Further, in the subject information fetching process 101, it is determined whether or not the division imaging examination is necessary based on the acquired subject information, and when the division imaging examination is necessary, the division number, that is, the set number of division examination areas is calculated, The set number is stored in the subject information storage unit 56 together with the number of cells in each divided examination region. As described above, the subject information acquired from the server 7 includes subject attribute information such as the width, height, number of constituent cells, and cell shape of the subject 1, examination item information such as examination range and pass / fail judgment value, Inspection condition parameter information such as coefficients. The determination of necessity of the division imaging inspection is performed by obtaining the size of the cell image in the inspection image based on the object information, particularly the object attribute information, and comparing it with a preset cell image threshold value. To do. More specifically, when the entire end face on the photographing side is captured in the inspection image and the size of the cell image in the inspection image is smaller than the cell image threshold value, the division imaging inspection is required. Then, the divided inspection area is set by dividing the photographing side end face. The number of divisions on the photographing side end face, that is, the size of the divided inspection area is set so that the size of the cell image in the inspection image of the divided inspection area unit is larger than the cell image threshold value.

  In the inspection image capturing process 102 subsequent to the object information capturing process 101, the imaging side end surface of the subject 1 is imaged by the camera 3 and the inspection image (this is an image of the entire imaging side end surface is divided and set. The inspection image may be an image of each of the divided inspection regions), and the inspection image is stored in the inspection image storage unit 58. FIG. 4 shows an example of an inspection image. An inspection image 31 in FIG. 4 is an inspection image in the case of a divided imaging inspection in which the imaging-side end surface of the subject 1 is divided into four, and is an inspection image of the upper left divided inspection area of the subject 1. In the examination image 31, an image of the cell 11 appears, an image of the partition wall 12 surrounding the four circumferences of the cell 11 appears, and an image of the side wall 13 of the subject 1 appears. Here, with respect to the cell 11, a completely penetrating cell 11a, a partially penetrating cell 11b, a non-penetrating cell 11c, and a penetrating cell are defined. The completely penetrating cell 11a is a cell that is substantially not clogged and transmits the illumination light from the illuminating device 2 without being blocked, and all the pixels in the cell region (the region surrounded by the partition wall 12) have a certain level or more. The partially penetrating cell 11b is a cell through which a part of the illumination light from the illuminating device 2 is transmitted, although the partially penetrating cell 11b is clogged, but the clogging is not completely blocked. , Only a part of the pixels in the cell region is defined as a cell having a certain level of brightness, and the non-penetrating cell 11c is completely blocked by clogging, and the illumination light from the illumination device 2 is substantially Are defined as cells in which all of the pixels in the cell region have a brightness below a certain level, and the penetrating cell is a cell that is a complete penetrating cell 11a or a partially penetrating cell 11b, that is, from the illumination device 2. Shine Or transmitted without light is blocked, or the illumination light is defined as a cell which transmits a part. In FIG. 4, the image of the partially penetrating cell 11b is shown as a circular bright portion and a dark portion around it. However, as shown in FIG. 6 and the like, the entire cell may appear as a dimmed image.

  When the inspection image capturing process 102 is completed, if the subject information capturing process 101 requires the division imaging examination, the division examination area set as the division imaging examination is not subjected to imaging. Presence / absence is determined by camera movement determination processing 103. If it is determined in the camera movement determination process 103 that there is an unphotographed divided inspection area, the camera 3 is moved on the XY stage 4 to a position where the unphotographed divided inspection area is imaged. FIG. 5 shows an example of division inspection area setting in the case of four-division imaging inspection and an example of movement of the camera 3 in the camera movement processing 104. In the quadrant imaging examination, the subject 1 is divided into four, and four divided examination areas 41A to 41D are set, and examination images 31 (31a to 31d) are sequentially photographed for each of the divided examination areas 41A to 41D. The sequential shooting of the inspection images 31 is performed by moving the shooting center of the camera 3, that is, the center P (Pa to Pd) of the inspection image 31, in the order of Pa → Pb → Pc → Pd. In such divided shooting, there may be a problem that the cells 11 at the boundaries of the inspection images 31a to 31d are cut off and shot. In order to cope with this problem, the camera 3 is moved so as to obtain an inspection image 31 having an overlap OV larger than the width of the cell 11 between the adjacent inspection images 31. By doing in this way, the problem of intermittent shooting of the boundary cell can be effectively solved. Here, the camera movement process 104 is a process in which the inspection process control unit 51 simply outputs a movement position and an operation start control signal to the XY stage 4. Therefore, in the determination processing device 5, the camera movement process 104 performed by the inspection process control unit 51 and the partition extraction process 105 and subsequent processes performed by the inspection process unit 52 can be performed in parallel. This eliminates the need for a waiting time for the camera movement process 104 and makes the inspection process more efficient.

  In the partition extraction process 105 performed in response to the inspection image input by the inspection image capturing process 102, the partition 12 of the cell 11 is extracted from the inspection image 31. The process can be performed as generation of a partition extraction image. FIG. 6 shows an image of an example of a partition extraction image generation process performed by the partition extraction process 105. The density of each pixel in the inspection image 31 (expressed by a brightness level: a scale value of 256 gradations with 0 being minimum and 255 being maximum) is high in the region of the cell 11 penetrating, and Low in area. Using this density difference relationship, the partition wall 12 is extracted to generate a partition wall extraction image. In the partition extraction image generation processing, first, a local maximum filter image 32 is generated from the inspection image 31 using a known local maximum filter technique. More specifically, by repeating the local maximum filter process by the number of pixels corresponding to the cell width, the local maximum filter image 32 having the density of the entire cell 11 (completely penetrating cell 11a) is generated. Once the local maximum filter image 32 is obtained, a difference image 33 is then generated by subtraction of the local maximum filter image 32 and the inspection image 31. Finally, a partition extraction image 34 is generated from the difference image 33. Specifically, the difference image 33 uses the fact that the density of each pixel is the difference in density between the local maximum filter image 32 and the inspection image 31, and uses a binarization threshold value for extracting only the partition wall portion. The image 33 is binarized to generate a partition extraction image 34.

  In the penetration cell extraction process 106 subsequent to the partition wall extraction process 105, a penetration cell (complete penetration cell and partial penetration cell) is extracted using the partition wall extraction image 34 and the inspection image 31. FIG. 7 illustrates an example of the penetration cell extraction process 106. In the penetrating cell extraction process 106 in the illustrated example, first, the density of the partition extraction image 34 is inverted to generate a partition extraction inverted image 35. Next, a partition removal image 36 is generated by taking the logical product of the partition extraction inverted image 35 and the inspection image 31. Finally, a penetration cell extraction image 37 and a complete penetration cell extraction image 38 are generated from the partition wall removal image 36. The penetration cell extraction image 37 is an image composed of the complete penetration cell 11a and the partial penetration cell 11b, and the complete penetration cell extraction image 38 is an image composed of only the complete penetration cell 11a. The penetration cell extraction image 37 is generated by binarizing the partition wall removal image 36 with a penetration cell extraction threshold. On the other hand, the complete penetration cell extraction image 38 is generated by binarizing the partition wall removal image 36 with a threshold for complete penetration cell extraction.

  It is necessary to extract all of the completely penetrating cell 11a. Therefore, the complete penetration cell extraction threshold is determined so that all the complete penetration cells can be extracted. Such a complete penetration cell extraction threshold value can be determined from the known density value when the density of the complete penetration cell is constant and the density is known throughout the entire divided inspection region on the examination image. However, when the lens used in the camera 3 has normal lens characteristics, the amount of light received by the light receiving surface of the camera 3 for the light beam traveling directly from the illumination device 2 tends to decrease as the distance from the center of the lens increases. In general, depending on the imaging conditions, the brightness varies in the periphery of the subject image on the examination image. In consideration of this, an average density is calculated for the partition wall removal image 36 actually obtained from the subject 1 to be examined at that time, and a predetermined coefficient (this is assumed to be a coefficient Ka). ) To determine the threshold for extracting completely through cells. The coefficient Ka for this purpose is experimentally obtained in advance for each type of subject and is one piece of subject information stored in the server 7. If the number of partially penetrating cells is large in the partition wall removal image 36, the average density of the partition wall removal image 36 becomes low, and it may be impossible to determine the threshold value for extracting completely through cells. In order to deal with such a case, the lowest concentration in the complete penetrating cell is measured in advance for each type of subject, and this is stored in the server 7 as the lower limit value of the threshold for extracting the complete penetrating cell. I will.

  On the other hand, the partial through-cell 11b only needs to be able to extract those having an opening ratio (which corresponds to the clogging ratio) of a certain value or more. Therefore, the penetration cell extraction threshold is determined according to the inspection condition. For example, a penetration cell extraction threshold in the case of an inspection condition for extracting a partial through-cell with a numerical aperture of up to 70% and a penetration cell extraction in the case of an inspection condition for extracting a partial through-cell with a numerical aperture of up to 30% The thresholds for use are different. Such a penetration cell extraction threshold value is set based on the complete penetration cell extraction threshold value and the aperture ratio of the extraction target partial penetration cell. For example, when the complete penetration cell extraction threshold is 120 and the aperture ratio of the extraction target partial penetration cell is 70%, the penetration cell extraction threshold is 100. That is, the penetration cell extraction threshold value is obtained by multiplying the complete penetration cell extraction threshold value by a coefficient corresponding to the aperture ratio of the extraction target partial penetration cell (assuming this is assumed to be a coefficient Kb). Such a coefficient Kb is experimentally obtained in advance for each type of subject and is one piece of subject information stored in the server 7.

  As described above, in the through cell extraction process 106, the complete through cell extraction threshold or the through cell extraction threshold set based on the average density of the partition wall removal image 36 is used. In this way, the partition wall removal image 36 is generated, and the threshold values for complete penetration cell extraction and the penetration cell extraction threshold value are set more appropriately based on the average density obtained for the partition wall removal image 36. Therefore, it is possible to improve the accuracy of extraction of complete through cells and through cells. That is, the removal of the partition wall by the partition wall extraction process 105 contributes to improving the accuracy of extraction of the complete through cell or the through cell. This is apparent when compared with the case where the partition wall removal process is not performed. That is, when the average density is obtained for the inspection image 31 without performing the partition wall removal process, and the threshold value for complete penetration cell extraction or the penetration cell extraction threshold value is set based on the average density, the influence of the density of the partition wall image in the inspection image 31 is affected. As a result, it is difficult to optimize each of these threshold values, and as a result, there is a risk that the accuracy of extraction of a complete through cell or through cell will be reduced.

  The above penetrating cell extraction process 106 is performed for each divided examination area when a plurality of divided examination areas are set as the division imaging examination necessity in the subject information fetching process 101. Then, in the extraction end determination process 107, it is determined whether or not the through cell extraction process 106 has been completed for all of the plurality of divided inspection areas. Returning to the capture processing 102, the subsequent processing is repeated.

  When it is determined in the extraction end determination process 107 that the extraction is completed, a cell penetration rate measurement process 108 is performed. In the cell penetration rate measurement process 108, first, the number of penetration cells, specifically the number of complete penetration cells and partial penetration cells, is measured. The number of penetrating cells can be measured using a known labeling method. When the number of penetrating cells is measured by the labeling method, the number of completely penetrating cells is obtained as the number of white labels by applying the labeling method to each pixel of the completely penetrating cell extracted image 38 and counting the number of white labels. On the other hand, in the measurement of the number of partially penetrating cells, first, as in the case of measuring the number of completely penetrating cells, the number of white labels is counted by applying a labeling method to each pixel of the penetrating cell extraction image 37, and the white label The number of penetrating cells (complete penetrating cells and partial penetrating cells) is obtained as the number, and then the number of penetrating cells is subtracted from the number of penetrating cells to obtain the number of partial penetrating cells.

  Here, in the case of the divided imaging inspection, as described with reference to FIG. 5, the inspection images 31 (31a to 31d) for each of the divided inspection regions have an overlap OV that is equal to or larger than the width of the cell 11 between the adjacent inspection images 31. Is filmed in this way. Therefore, for example, the inspection image 31c and the inspection image 31d include a common cell image. In this case, if the common cell is measured as a complete penetration cell or a partial penetration cell in each of the inspection image 31c and the inspection image 31d, double measurement is performed. Such double measurement can be avoided by using the characteristics of the labeling method. That is, in the labeling method, region information (start point coordinates, end point coordinates, center coordinates, etc.) on the image of the extracted penetrating cells (complete penetrating cells and partial penetrating cells) can also be measured. Therefore, by adding the area information on the image as a condition for measuring the number of labels, it is possible to make a condition so that the common cell is measured only for one inspection image 31, and double measurement can be avoided. The area information obtained by the labeling method can be used in addition to avoiding such double measurement. For example, by expanding the area information of the penetrating cell measured for each divided inspection area in the divided photographing inspection into a unified coordinate system, it is possible to measure the penetrating cell as in the case of the divided photographing inspection. Further, the vicinity of the outer edge of the honeycomb structure 1 is also an area where the inspection object exclusion portion is masked with the coordinates on the image when it is desired to exclude the inspection object from the inspection object because it is likely to be clogged. Information can be used.

  When the measurement of the number of penetrating cells is completed as described above, the cell penetrating rate is obtained next. As the cell penetration rate, each of the penetration cell rate, the complete penetration cell rate, the partial penetration cell rate, and the penetration cell component ratio is obtained. For the penetration cell rate, the number of penetration cells (the number of complete penetration cells + the number of partial penetration cells) is held in the subject information storage unit 56 as the number of constituent cells of the subject 1 (this is subject information acquired from the server 7). ), And the complete penetration cell rate is calculated by dividing the number of complete penetration cells by the number of constituent cells of the subject 1. The partial penetration cell rate is calculated by dividing the number of partial penetration cells by the number of the subject 1. The through cell configuration ratio is calculated by dividing the number of through cells by the number of through cells. The cell penetration rate data obtained in this way is stored in the inspection result storage unit 57.

  When the cell penetration rate is obtained in the cell penetration rate measurement process 108, a pass / fail judgment process 109 is performed based on the cell penetration rate. In the pass / fail determination process 109, pass / fail determination under various conditions is performed by comparing the cell penetration rate (penetration cell rate, complete penetration cell rate, partial penetration cell rate, penetration cell configuration ratio) with preset determination values. Can do. For example, by combining the cell penetration rate and the penetration cell composition ratio, the subject 1 having a high cell penetration ratio but a low penetration cell composition ratio or a subject 1 having a high penetration cell composition ratio but a low cell penetration ratio is regarded as defective. Advanced pass / fail judgment can be made. In addition, for example, in the case of an inspection condition in which the vicinity of the outer edge of the subject 1 is lightly considered to be less affected by clogging, it is possible to perform pass / fail determination by weighting the penetrating cell and lowering the determination criterion of the outer edge portion from the center portion. It is.

  According to the inspection method as described above, the cell penetration rate (through cell rate, complete penetration cell rate, partial penetration cell rate, penetration) obtained by individually measuring the penetration cells (complete penetration cell and partial penetration cell) in the subject. Since the quality of the subject is determined based on the cell composition ratio), the influence of the partition walls of the cells can be eliminated, and high-accuracy determination regarding clogging can be performed stably.

  Here, various data obtained by the above examination processing can be effectively used in addition to the quality determination of the subject 1. For example, by statistically processing the cell penetration rate for a plurality of specimens, it is possible to quickly grasp the defect occurrence tendency in the manufacturing process of the honeycomb structure. Further, for example, it is possible to estimate the cause of clogging from the through-cell composition ratio. In this respect, the above inspection method is more effective.

It is a figure which shows typically the hardware constitutions of the honeycomb structure inspection system by one Embodiment. It is a figure which shows the structural example of the determination processing apparatus. It is a figure which shows the flow of the process in the inspection process by the honeycomb structure inspection system of FIG. It is a figure which shows the example of a test | inspection image. It is a figure which shows the example of a division | segmentation test | inspection area | region setting, and the example of a camera movement. It is a figure which shows an example of a partition extraction image generation process as an image. It is a figure which imageizes and shows an example of a penetration cell extraction process.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Honeycomb structure 2 Illuminating device 3 Camera 5 Judgment processing apparatus 11 Cell 11a Complete penetration cell (penetration cell)
11b Partial penetration cell (penetration cell)
31 Inspection image 36 Bulkhead removal image

Claims (8)

In order to inspect the clogging of the cells in the honeycomb structure in which a large number of cells are formed in a honeycomb shape, the honeycomb structure is illuminated from one end side so that illumination light is transmitted through the cells by an illumination device. In the inspection method in which the end face on the other end side of the honeycomb structure is photographed to obtain an inspection image, and whether or not the cell clogging is determined based on the inspection image.
The number of penetrating cells that pass through the illumination light from the illuminating device without being obstructed or through which the illuminating light from the illuminating device is partially transmitted are measured from the inspection image, and the number of the penetrating cells. The inspection method is characterized in that the pass / fail judgment is performed based on the above.   2. The pass / fail ratio is obtained by dividing the number of penetrating cells by the total number of cells in the honeycomb structure, and the quality determination is performed based on the penetrating cell ratio. Inspection method.   The measurement of the number of through-cells is performed based on a partition removal image generated by removing an image of the partition of the cell from the inspection image. Inspection method.   By measuring the number of completely penetrating cells, which are cells through which the illumination light from the lighting device is transmitted without being blocked, from the inspection image, and dividing the number of completely penetrating cells by the total number of cells in the honeycomb structure. The inspection method according to any one of claims 1 to 3, wherein a complete penetration cell rate is obtained, and the complete penetration cell rate can be used for the quality determination.   5. The penetration cell composition ratio is obtained by dividing the number of complete penetration cells by the number of penetration cells, and the penetration cell composition ratio can be used for the pass / fail judgment. Inspection method.   The number of partially penetrating cells, which are cells through which the illumination light from the lighting device is partially transmitted, is measured from the inspection image, and the partial penetrating cell number is divided by the total number of cells in the honeycomb structure. The inspection method according to any one of claims 1 to 5, wherein a cell rate is obtained, and the partially penetrating cell rate can be used for the quality determination.   The end face of the honeycomb structure is divided into a plurality of divided inspection regions, the inspection image is acquired for each of the divided inspection regions, and the quality determination can be performed based on the plurality of inspection images. The inspection method according to any one of claims 1 to 6. An illumination device that illuminates the honeycomb structure from one end side so that illumination light is transmitted through the cells in order to inspect the clogging of the cells in the honeycomb structure in which a large number of cells are formed in a honeycomb shape. In an inspection system comprising a camera for obtaining an inspection image by photographing the end face on the other end side of the honeycomb structure in a state illuminated with the illumination device,
The number of penetrating cells that pass through the illumination light from the illuminating device without being obstructed or through which the illuminating light from the illuminating device is partially transmitted are measured from the inspection image, and the number of the penetrating cells. An inspection system comprising: a determination processing device that performs the pass / fail determination based on the inspection method, wherein the inspection method according to any one of claims 1 to 7 can be executed.

Images