JP2008139500A - Image analyzing device, fault diagnosing device, image analyzing program, and fault diagnosing program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To precisely detect a period even for image data having an image feature of indistinct periodicity. <P>SOLUTION: An image analyzing device includes a difference image data calculating means 12 for calculating difference image data between read image data obtained by forming an image on the basis of a predetermined test chart by an image forming apparatus and then reading the image by an image read unit and original data of the test chart; a projection waveform calculating means 14 for calculating projection waveform data by adding the difference image data to a main scanning direction; a defect interval detecting means 16 for detecting occurrence intervals of an image defect on the basis of a position of a prescribed value about a peak value of a rise of the projection waveform data as an edge position of the image defect; and a defect period detecting means 18 for calculating difference distances for a plurality of reference values as to the detected defect intervals, calculating defect period candidate probabilities for the reference values on the basis of the calculated difference distances, and outputting a defect period having the highest calculated period candidate probability to a monitor such as a console panel. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to an image analysis device, a failure diagnosis device, an image analysis program, and a failure diagnosis program.

  Analysis of image data having image characteristics including at least periodicity and detection of periodicity perform some processing or use as reference information when performing processing.

  For example, an image forming apparatus such as an electrophotographic copying machine or a printer is configured to include a large number of image forming members made of a rotating body. In the failure diagnosis for identifying a defective portion that causes an image defect, The image defect period (pitch) based on the peripheral length of the image forming member is effective diagnosis information, and can be used as failure diagnosis information by detecting the period of the image defect.

  As techniques for analyzing image defects, for example, techniques described in Patent Document 1 and Patent Document 2 have been proposed.

  In the technique described in Patent Document 1, in a line interval evaluation device that evaluates unevenness in intervals between a plurality of lines having a small interval, the brightness of an image photographed by a television camera is added in the direction of a line pattern that is an image of the line. By calculating a projection distribution which is a distribution perpendicular to the line pattern of the value obtained above, determining the position of the center line of each concave portion of the projection distribution, and determining the interval between adjacent center lines, a plurality of line intervals It has been proposed to calculate the amount of variation.

Further, in the technique described in Patent Document 2, image data of a region of interest selected and designated from among image regions of image data having a plurality of periodic patterns is set as processing target image data, and quantification is performed on the processing target image data. When a pattern edge is detected by performing typical image processing, a representative pattern that is a representative pattern based on a predetermined rule from among a plurality of edge candidates having predetermined image characteristics is processed for image data to be processed It has been proposed to perform a representative edge detection process for detecting an edge as a first processing step of the edge detection process. Thereby, an image with poor edge quality can be quantified, and an edge can be detected with high accuracy.
JP-A-3-282684 JP 2004-272313 A

  The present invention has been made in the background of the background art as described above, and is an image analysis apparatus and a failure diagnosis which can detect a period with high accuracy even for image data having an image feature whose periodicity is unclear. An object is to provide an apparatus, an image analysis program, and a failure diagnosis program.

  In order to achieve the above object, the image analysis apparatus according to claim 1, a projection waveform calculation unit that calculates a projection waveform in a direction that intersects a periodic direction of periodicity of image data having image characteristics including at least periodicity; Periodic interval detecting means for detecting a periodic interval included in the image data based on a predetermined edge position of the projected waveform calculated by the projected waveform calculating means.

  The image analysis apparatus according to claim 2, wherein a projection waveform calculation unit that calculates a projection waveform in a direction that intersects a periodic direction of periodicity of image data having image characteristics including at least periodicity, and the projection waveform calculation unit calculates And a periodic interval detecting means for obtaining a self-correlation of the projected waveform and detecting a periodic interval included in the image data based on the correlation.

  The image analysis apparatus according to claim 3, a difference calculation unit that calculates difference image data between image data having image characteristics including at least periodicity and predetermined image data, and the difference calculated by the difference calculation unit Projection waveform calculation means for calculating a projection waveform in a direction intersecting with a periodic direction of periodicity of the image data from image data, and the image data based on a predetermined edge position of the projection waveform calculated by the projection waveform calculation means And a periodic interval detecting means for detecting a periodic interval included in the periodicity.

  The image analysis apparatus according to claim 4, a difference calculation unit that calculates difference image data between image data having image characteristics including at least periodicity and predetermined image data, and a difference calculated by the difference calculation unit A projection waveform calculation unit that calculates a projection waveform in a direction that intersects with a periodic direction of the periodicity of the image data from image data, and obtains a correlation between the projection waveform calculated by the projection waveform calculation unit and the correlation And a periodic interval detecting means for detecting periodic periodic intervals included in the image data.

  An image analysis apparatus according to a fifth aspect of the present invention is the image forming apparatus according to any one of the first to fourth aspects, wherein the periodic interval detected by the periodic interval detecting means and an image are formed. A candidate probability for each reference value of periodicity, which is an image feature included in the image data, based on a plurality of predetermined reference values representing the period of a component that causes a periodic image in The image processing apparatus further includes detection means for detecting a periodic period included in the image data based on the calculated candidate probability.

  According to a sixth aspect of the present invention, in the invention according to any one of the first to fifth aspects, the projection waveform calculating means is configured to detect a direction intersecting with a periodic direction of the image data. Then, the projection waveform is calculated by performing addition processing.

  According to a seventh aspect of the present invention, in the image analysis apparatus according to any one of the first to sixth aspects, the projected waveform calculating means calculates a moving average of the calculated projected waveform. It is characterized by including.

  The image analysis device according to claim 8 is the image analysis device according to any one of claims 1 to 7, wherein the image data is calculated when the projection waveform calculation unit calculates the projection waveform. A binarizing means for binarizing is further provided.

  The image analysis apparatus according to claim 9 is the invention according to any one of claims 5 to 8, wherein the detection unit corresponds to the highest candidate probability among the calculated candidate probabilities. The reference value is detected as a periodic period included in the image data.

  An image analysis apparatus according to a tenth aspect is the invention according to any one of the fifth to eighth aspects, wherein the detection means uses a plurality of reference values corresponding to the calculated candidate probabilities. It is characterized in that it is detected as a periodicity period included in image data.

  An image analysis apparatus according to an eleventh aspect of the invention is characterized in that in the invention according to any one of the first to tenth aspects, the image analysis apparatus further comprises a notifying means for notifying a detection result of the detecting means. .

  The failure diagnosis apparatus according to claim 12, a difference calculation unit that calculates difference image data between image data having image characteristics including at least periodicity and predetermined image data, and a difference calculated by the difference calculation unit Projection waveform calculation means for calculating a projection waveform in the periodic direction of the periodicity of the image data from the image data, and included in the image data based on a predetermined edge position of the projection waveform calculated by the projection waveform calculation means Periodic interval detecting means for detecting a periodic interval of periodicity, a periodic interval detected by the periodic interval detecting means, and a period of a component that causes a periodic image in an image forming apparatus that forms an image are predetermined. A candidate for calculating a candidate probability for each reference value of periodicity, which is an image feature included in the image data, based on a plurality of reference values A rate calculating means, is characterized in that and a fault diagnosis means for the candidate probability that is calculated by inputting a probabilistic model performs failure diagnosis of the image forming apparatus by the candidate probability calculation means.

  The fault diagnosis apparatus according to claim 13, a difference calculation unit that calculates difference image data between image data having image characteristics including at least periodicity and predetermined image data, and a difference calculated by the difference calculation unit A projection waveform calculation unit that calculates a projection waveform in a direction that intersects with a periodic direction of the periodicity of the image data from image data, and obtains a correlation between the projection waveform calculated by the projection waveform calculation unit and the correlation Based on the periodic interval detection means for detecting the periodic interval included in the image data, the periodic interval detected by the periodic interval detection means, and the cause of the periodic image in the image forming apparatus for forming an image. Based on a plurality of predetermined reference values representing the period of the constituent parts, and each reference value of periodicity that is an image feature included in the image data. A candidate probability calculation unit that calculates a candidate probability, and a failure diagnosis unit that inputs the candidate probability calculated by the candidate probability calculation unit into a probability model and performs a failure diagnosis of the image forming apparatus. Yes.

  The image analysis program according to claim 14 is calculated by a projection waveform calculation step of calculating a projection waveform in a direction intersecting a periodic direction of periodicity of image data having image characteristics including at least periodicity, and the projection waveform calculation step. And a periodic interval detecting step of detecting a periodic interval included in the image data based on a predetermined edge position of the projected waveform thus obtained.

  The image analysis program according to claim 15 is calculated by a projection waveform calculation step of calculating a projection waveform in a direction intersecting a periodic direction of periodicity of image data having image characteristics including at least periodicity, and the projection waveform calculation step. And a periodic interval detecting step of obtaining a periodic interval of periodicity included in the image data on the basis of the correlation of the projected waveform thus obtained, and causing the computer to execute processing including: It is said.

  The image analysis program according to claim 16, wherein a difference calculation step of calculating difference image data between image data having image characteristics including at least periodicity and predetermined image data, and the difference calculated by the difference calculation step A projection waveform calculation step for calculating a projection waveform in a direction intersecting with a periodic direction of periodicity of the image data from the image data, and the image data based on a predetermined edge position of the projection waveform calculated by the projection waveform calculation step And a periodic interval detecting step for detecting a periodic interval included in the computer.

  The image analysis program according to claim 17, wherein a difference calculation step of calculating difference image data between image data having image characteristics including at least periodicity and predetermined image data, and the difference calculated by the difference calculation step A projection waveform calculation step for calculating a projection waveform in a direction intersecting with the periodic direction of the periodicity of the image data from the image data, and obtaining a correlation between the projection waveform calculated by the projection waveform calculation step, and calculating the correlation On the basis of this, a computer is caused to execute a process including a periodic interval detecting step of detecting a periodic interval included in the image data.

  The fault diagnosis program according to claim 18, wherein a difference calculation step of calculating difference image data between image data having image characteristics including at least periodicity and predetermined image data, and the difference calculated by the difference calculation step A projection waveform calculation step for calculating a projection waveform in a direction intersecting with a periodic direction of periodicity of the image data from the image data, and the image data based on a predetermined edge position of the projection waveform calculated by the projection waveform calculation step A periodic interval detecting step for detecting a periodic interval included in the periodic interval, a periodic interval detected by the periodic interval detecting step, and a period of a component that causes a periodic image in an image forming apparatus that forms an image. A period that is an image feature included in the image data based on a plurality of predetermined reference values A candidate probability calculating step of calculating a candidate probability for each of the reference values, and a failure diagnosis step of performing a failure diagnosis of the image forming apparatus by inputting the candidate probability calculated by the candidate probability calculating step into a probability model. It is characterized by causing a computer to execute processing.

  The fault diagnosis program according to claim 19, wherein a difference calculation step of calculating difference image data between image data having image characteristics including at least periodicity and predetermined image data, and the difference calculated by the difference calculation step A projection waveform calculation step for calculating a projection waveform in a direction intersecting with the periodic direction of the periodicity of the image data from the image data, and obtaining a correlation between the projection waveform calculated by the projection waveform calculation step, and calculating the correlation A periodic interval detection step for detecting a periodic interval included in the image data, a periodic interval detected by the periodic interval detection step, and a cause of the periodic image in the image forming apparatus for forming an image. An image included in the image data based on a plurality of predetermined reference values representing the cycle of the component A candidate probability calculating step for calculating a candidate probability for each reference value of periodicity, which is a characteristic, and a failure diagnosis for inputting the candidate probability calculated by the candidate probability calculating step into a probability model to perform a failure diagnosis of the image forming apparatus And causing the computer to execute a process including the steps.

  According to the image analysis apparatus of the first aspect, since the projection waveform is calculated based on the direction having periodicity, the period of the periodic image can be detected with higher accuracy than when the configuration is not provided. can do.

  According to the image analysis apparatus of the second aspect, since stability against uncertain fluctuations such as disturbances and errors can be improved, the periodic image is compared with the case where the present configuration is not provided. The period can be detected.

  According to the image analysis apparatus of the third aspect, since the projection waveform is calculated based on the direction having periodicity from the difference image data with the predetermined image data, it is compared with the case where this configuration is not provided. Thus, the period of the periodic image can be detected with high accuracy.

  According to the image analysis apparatus of the fourth aspect, the stability with respect to uncertain fluctuations such as disturbances and errors can be improved from the difference image data with the predetermined image data. The period of the periodic image can be detected as compared with the case where there is not.

  According to the image analysis apparatus of the fifth aspect, since the periodicity period included in the image data is detected based on the candidate probability with respect to each reference value of periodicity, it is compared with the case where this configuration is not provided. Thus, the period of the periodic image can be detected with high accuracy.

  According to the image analysis apparatus of the sixth aspect, since the projection waveform is calculated by performing the addition process on the direction intersecting the surrounding direction, the periodicity can be emphasized, so this configuration is provided. Compared with the case where it is not, the period of a periodic image can be detected with high accuracy.

  According to the image analysis apparatus of the seventh aspect, since the moving average process is performed on the projection waveform, it is possible to reduce erroneous detection of the period of the periodic image compared to the case where the present configuration is not provided. Can do.

  According to the image analysis apparatus of the eighth aspect, since the binarization process is performed when calculating the projection waveform, the noise is removed and the periodic image is compared with the case where the present configuration is not provided. The period can be detected accurately.

  According to the image analysis device of claim 9, since the reference value corresponding to the highest candidate probability is detected as the periodicity period included in the image data, compared with the case where this configuration is not provided, It is possible to reliably identify the components of the image forming apparatus due to the period of the periodic image.

  According to the image analysis device of claim 10, since a plurality of reference values corresponding to each candidate probability are detected as a periodicity period included in the image data together with the candidate probabilities, this configuration is not provided. Compared to the above, it is possible to obtain information for specifying the components of the image forming apparatus due to the period of the periodic image.

  According to the image analysis apparatus of the eleventh aspect, since the detection result is notified, the detected periodicity period can be confirmed by display or the like as compared with the case where the present configuration is not provided.

  According to the failure diagnosis apparatus according to claim 12, the present embodiment does not have this configuration because the candidate probability of the period of the periodic image is accurately calculated and input to the probability model to perform the failure diagnosis of the image forming apparatus. Compared to the case, the accuracy of the fault diagnosis of the image forming apparatus can be improved.

  According to the failure diagnosis apparatus according to claim 13, since the candidate probability with improved stability against uncertain fluctuations such as disturbance and error is calculated and input to the probability model, failure diagnosis of the image forming apparatus is performed. Compared to the case where this configuration is not provided, the accuracy of failure diagnosis of the image forming apparatus can be improved.

  According to the image analysis program of claim 14, since the projection waveform is calculated based on the direction having periodicity, the period of the periodic image can be detected with higher accuracy than in the case where the present configuration is not provided. can do.

  According to the image analysis program of the fifteenth aspect, stability against uncertain fluctuations such as disturbances and errors can be improved. Therefore, compared with the case where the present configuration is not provided, the periodic image The period can be detected.

  According to the image analysis program of the sixteenth aspect, the projection waveform is calculated based on the direction having periodicity from the difference image data with the predetermined image data, so that it is compared with the case where the present configuration is not provided. Thus, the period of the periodic image can be detected with high accuracy.

  According to the image analysis program of the seventeenth aspect, the stability with respect to uncertain fluctuations such as disturbances and errors can be improved from the difference image data with the predetermined image data. The period of the periodic image can be detected as compared with the case where there is not.

  According to the failure diagnosis program of claim 18, the present embodiment does not have this configuration because the candidate probability of the period of the periodic image is accurately calculated and input to the probability model to perform the failure diagnosis of the image forming apparatus. Compared to the case, the accuracy of the fault diagnosis of the image forming apparatus can be improved.

  According to the failure diagnosis program according to claim 19, since the candidate probability with improved stability against uncertain fluctuations such as disturbances and errors is calculated and input to the probability model, the image forming apparatus is diagnosed for failure. Compared to the case where this configuration is not provided, the accuracy of failure diagnosis of the image forming apparatus can be improved.

  Hereinafter, an example of an embodiment of the present invention will be described in detail with reference to the drawings. FIG. 1 is a diagram showing a schematic configuration of an image forming apparatus for recording an image for detecting an image defect by the image analysis apparatus according to the embodiment of the present invention.

  The image forming apparatus 50 forms an image using an electrophotographic system, and mainly includes an image output unit 52, an image reading unit 54, and an operation panel 56 as shown in FIG.

  The operation panel 56 serves as a user interface between the image forming apparatus 50 and the user, such as an operation instruction input of the image forming apparatus 50, device information display, and operation for failure diagnosis, and includes input devices such as a display monitor and a keyboard.

  The image output unit 52 applies an example of an image recording device using an optical scanning device, and includes four image recording units corresponding to K (black), Y (yellow), M (magenta), and C (cyan) colors. A so-called tandem type image output unit is provided. The image output unit 52 includes image forming units 58 (58K, 58Y, 58M, and 58C) for each color, which are juxtaposed sequentially at a predetermined interval in one direction, and an intermediate transfer belt 60. Note that symbols K, Y, M, and C in FIG. 1 are described corresponding to the colors K, Y, M, and C, and in the following description, symbols that correspond to the colors when the colors are described separately. However, when the colors are not distinguished, the symbols corresponding to the colors may be omitted for explanation.

  In addition, the image forming unit 58 includes an optical scanning device having rotating polygon mirrors 64YM and 64CK that reflect laser light emitted from a semiconductor laser (not shown) toward a photosensitive drum 62 that is an example of a photosensitive member. Yes.

  Below the image output unit 52, there are a paper feeding unit 66 provided with paper for feeding to the image forming unit 58, and a paper conveyance path for conveying the paper of the paper feeding unit 66 to the intermediate transfer belt 60. Is provided. Note that the paper transport path includes a pick-up roll 68 and a plurality of transport rolls 70 for taking out the paper of the paper feed unit 66 one by one.

  For example, in the black (K) image forming unit 58K, first, the semiconductor laser is driven by a black image forming signal from an image processing unit (not shown) to convert the black image forming signal into an optical signal, The converted laser light is irradiated toward the polygon mirror 64CK. The laser light further scans the photosensitive drum 62K charged by the charging roll 74K via the reflection mirror 72, thereby forming an electrostatic latent image on the photosensitive drum 62K. The electrostatic latent image is converted into a toner image by a developing device 76K to which black toner is supplied. The toner image is transferred by an intermediate transfer belt 78K by a primary transfer roll 78K while the intermediate transfer belt 60 passes through the photosensitive drum 62. 60 is transferred. After the transfer, excess toner is removed from the photosensitive drum 62 by the cleaner 80K. Similarly, electrostatic latent images are formed on the photosensitive drums 62Y, 62M, and 62C by corresponding Y, M, and C image forming signals sequentially obtained from the image processing unit at predetermined intervals with respect to the black image forming signals. An image is formed, and each electrostatic latent image is sequentially converted into a toner image by developing units 76Y, 76M, and 76C to which toner of each color is supplied. Each toner image is a photosensitive drum 62Y, corresponding to the intermediate transfer belt 60. The images are sequentially transferred onto the intermediate transfer belt 60 by the corresponding primary transfer rolls 78M, 78Y, 78C while passing through 62M, 62C.

  The toner on the intermediate transfer belt 60 on which the toner images of each color of K, Y, M, and C are successively transferred in multiple layers is peeled off from the intermediate transfer belt 60 by the secondary transfer roll 82, and the toner is fixed by the fixing device 84. And discharged outside the copier. A discharge path for discharging the sheet to the outside of the apparatus is provided on the downstream side of the fixing device 84 in the sheet conveyance direction, and the image output unit 52 receives the printed sheet formed on the printing sheet outside the apparatus. A paper discharge unit 86 is also provided.

  The image reading unit 54 according to the present embodiment optically reads an image drawn on a document from a sheet-like document to be read, and includes a platen cover 88 that covers the document. The image reading unit 54 has a platen glass 90 on which a document to be read is placed, and an optical system including a light receiving unit 92 for reading the document and an image processing unit on the image reading unit side below the platen glass 90. ing. The image reading unit 54 includes a light source 94 that emits light toward a surface opposite to the surface of the platen glass 90 on which the document is placed, below the platen glass 90, and emits light emitted from the light source 94. A full-rate carriage having a reflective shade that reflects to the platen glass 90 side, a reflective mirror, and a reflective mirror 96 that deflects reflected light from the platen glass 90 side in a direction substantially parallel to the platen glass 90 is provided.

  As the light source 94, a fluorescent lamp whose longitudinal direction is the main scanning direction (the direction orthogonal to the drawing in the drawing) can be applied. As the color of the illumination light of the light source 94, a color that matches the spectral optical characteristics of each line sensor that constitutes the light receiving unit 92 is used. For example, white light or green light is used. In addition, the image reading unit 54 includes a lens 98 that condenses the reflected light deflected by the reflection mirror 96 of the half-rate carriage at a predetermined focal position. The light receiving unit 92 receives the reflected light converged by the lens 98, reads an image in the main scanning direction substantially orthogonal to the sub-scanning direction, and sequentially outputs image signals corresponding to the density. The image signal from the image reading unit 54 is sequentially output to the document data storage unit via a read image processing unit (not shown). The document data stored in the document data storage unit is the operation instruction of the image forming apparatus input from the operation panel 56 (multiple pages, booklet, multi-page composition to make N pages into one page, double-sided, booklet double-sided, enlarged continuous shooting) And the like are sequentially output to the image output unit 52.

  Among the image defects in the image forming apparatus of the present embodiment, those that cause periodic image defects include, for example, the photosensitive drum 62, the charging roll 74, the developing roll in the developing device 76, and the primary transfer roll 78. Secondary transfer roll 82, heat roll in fixing device 84, intermediate transfer belt 60, and the like. Examples of the image defect period (pitch) caused by each image forming member include, for example, the outer periphery of the photosensitive drum, the outer periphery of the charging roll, the outer periphery of the developing roll, the outer periphery of the primary transfer roll, the outer periphery of the secondary transfer roll, the outer periphery of the heat roll, and the intermediate transfer belt. There are an outer periphery, a photosensitive drum interval, and the like. Examples of each image defect period (pitch) in a small electrophotographic image forming apparatus include a photoreceptor drum outer periphery 99.5 mm, a charging roll outer periphery 46.6 mm, a developing roll outer periphery 29.6 mm, a primary transfer roll outer periphery 61.4 mm, There are cases where the outer periphery of the secondary transfer roll is 93.2 mm, the outer periphery of the heat roll is 88.9 mm, the outer periphery of the intermediate transfer belt is 1005.2 mm, and the photosensitive drum interval is 95.7 mm. In the above example, the values of the heat roll outer periphery 88.9 mm, the secondary transfer roll outer periphery 93.2 mm, the photosensitive drum interval 95.7 mm, and the photosensitive drum outer periphery 99.5 mm are close to each other, and an accurate defect period ( Pitch) is essential. Furthermore, twice the outer circumference of the charging roll 46.6 mm corresponds to the secondary transfer roll outer circumference 93.2 mm, and erroneous detection of the defect cycle (pitch) leads to erroneous diagnosis.

  Subsequently, an image analysis apparatus according to an embodiment of the present invention that detects an image defect formed on a sheet-like document by the above-described image forming apparatus will be described in detail.

(First embodiment)
FIG. 2 is a block diagram showing a schematic configuration of the image analysis apparatus 10 according to the first embodiment of the present invention. The image analysis apparatus 10 may be mounted on the image forming apparatus 50 or may be configured separately, but in the present embodiment, the image analysis apparatus 10 will be described as mounted on the image forming apparatus 50.

  As shown in FIG. 2, the image analysis apparatus 10 according to the first embodiment of the present invention includes difference image data calculation means 12, projection waveform calculation means 14, defect interval detection means 16, and defect period detection means 18. Yes.

  The difference image data calculation unit 12 forms an image based on a predetermined test chart by the image forming apparatus 50 and reads the read image data obtained by reading by the image reading unit 54 and the original data of the test chart (test chart image data). ), And difference processing is obtained. As the test chart, a test chart corresponding to an image defect is prepared in advance. It is desirable to use a blank test chart, a halftone gray chart, or the like so that image defects can be seen. For example, FIG. 16A shows an example of a blank test chart, and an example of a test chart including a periodic image defect (an example of an image defect of the charging roll outer periphery period) with respect to the blank test chart is shown in FIG. Shown in B). FIG. 17 shows a difference image between FIG. 16 (A) and FIG. 16 (B). In each figure, the horizontal direction is the main scanning direction, and the vertical direction is the sub-scanning direction.

  The projection waveform calculation means 14 performs addition processing with respect to the main scanning direction of difference image data. That is, the projection waveform data is data obtained by projecting difference image data in the main scanning direction by adding data in the main scanning direction. By performing the addition process in the main scanning direction in this way, the periodicity of low-density unclear streaky irregularities and strip-shaped irregularities that occur periodically in the sub-scanning direction is emphasized. For example, the projection waveform data is as shown in FIGS. 3A is a diagram illustrating an example of a projection waveform of an image defect in the outer peripheral cycle of the photosensitive drum 62, and FIG. 3B is a diagram illustrating an example of a projection waveform of an image defect in the outer peripheral cycle of the charging roll 74. It is. 3A and 3B, the horizontal axis is a data point corresponding to an A4 size of 210 mm with a resolution of about 90 μm, and the vertical axis is a value obtained by normalizing the added value in the main scanning direction of the difference image data. It is.

  The defect interval detection unit 16 calculates the interval at which defects occur from the projection waveform data calculated by the projection waveform calculation unit 14. In the present embodiment, with respect to the rising edge of the projection waveform data, the image defect occurrence interval is set as the edge position of the image defect with reference to a position that is a predetermined value with respect to the peak value (in this embodiment, half the peak value). To detect.

  However, as can be seen from the example of FIG. 3A, a cluster of one image defect may have a plurality of peaks and may be equal to or less than half the peak value at the center. In this case, defect interval detection means When the image defect interval is detected by 16, erroneous detection occurs. Therefore, as shown in FIG. 4, the projection waveform calculation means 14 may include a projection waveform calculation processing unit 14A and a moving average processing unit 14B. That is, after the projection waveform data is generated by the projection waveform calculation processing unit 14A, the moving average processing unit 14B can apply the moving average processing to the projection waveform data. FIGS. 5A and 5B show an example of projection waveform data when the averaging process is performed for every 75 points with respect to FIGS. 3A and 3B. The defect interval when the above algorithm is applied to the projection waveform data of FIG. 5A is 101.9 mm. Similarly, the defect interval when the above algorithm is applied to the projection waveform data of FIG. Are 47.5 mm, 45.2 mm, and an average of 46.4 mm. Note that 47.5 mm and 45.2 mm respectively correspond to the peak defect interval of the second and third projection waveform data in FIG. 5B and the peak defect interval of the third and fourth projection waveform data. . The values are close to the outer periphery of the photosensitive drum 62 of 99.5 mm and the outer periphery of the charging roll 74 of 46.6 mm. In the above algorithm, the rising edge of the projection waveform is used as a calculation reference for the defect occurrence interval. However, the present invention is not limited to this. For example, the falling edge of the projection waveform data may be used.

  Note that when the predetermined value for the peak value when the image defect interval is detected by the defect interval detector 16 is set small, the error increases due to the influence of the occurrence state of the image defect peripheral portion. In the present embodiment, as an example, the half value is set to a predetermined value, but is not limited to the half value, and can be set as appropriate according to, for example, the type of image defect.

  The defect period detection unit 18 corrects the defect interval detected by the defect interval detection unit 16 based on the periodicity (pitch) originally generated in the image forming member. Specifically, a plurality of reference values (for example, the photosensitive drum 62, the charging roll 74, the developing roll, the primary transfer roll 78, the secondary transfer roll 82, the heat roll, and the intermediate transfer belt) with respect to the detected defect interval. A reference distance such as 60, which is stored in advance in a memory or the like), and a defect distance candidate probability (a candidate probability for the reference value 20) is calculated based on the calculated difference distance; The calculated defect period with the highest defect period candidate probability is output to a monitor such as the operation panel 56. The defect cycle candidate probability may be simply proportional to the distance from the reference position, but it is desirable to set the defect cycle candidate probability based on the defect cycle distribution generated due to the individual image forming members. The calculation of the defect periodicity probability is the average value of the periodicity (pitch) of each image forming member, and the standard deviation calculated from the defect cycle distribution generated due to each image forming member is the standard deviation. Assuming a normal distribution, the two-sided probability for the difference distance difference is calculated.

  For example, the defect interval detected based on FIG. 5A is 101.9 mm, but when an image forming member having a period of ± 10 mm of 101.9 mm is extracted, the outer periphery of the photosensitive drum 62 is 99.5 mm, the photosensitive drum 62 distance 95.7 mm and the outer periphery 93.2 mm of the secondary transfer roll 82 are extracted, and the difference distance is the outer periphery 2.4 mm of the photosensitive drum 62, the photosensitive drum 62 interval 6.2 mm, and the outer periphery of the secondary transfer roll 82 8.7 mm. . Further, assuming that the standard deviation of the defect period of each image forming member is 4.2 mm on the outer periphery of the photosensitive drum 62, 3.4 mm between the photosensitive drums 62, and 5.1 mm on the outer periphery of the secondary transfer roll 82, the respective defect period candidate probabilities are Photoconductor drum 62 outer periphery 57% (2.4 mm / 4.2 mm = 0.57 both-sides probability), photoconductor drum 62 interval 7% (6.2 mm / 3.4 mm = 1.82 both-sides probability), secondary transfer roll 82 outer periphery 9% It is determined that the probability of the outer periphery of the photosensitive drum 62 is 99.5 mm is the highest, and a defect period of 99.5 mm is output.

  Similarly, the average defect interval detected based on FIG. 5B is 46.4 mm, but when an image forming member having a period of ± 10 mm of 46.4 mm is extracted, only the outer periphery of the charging roll 74 is 46.6 mm. If the standard deviation of the defect period on the outer periphery of the charging roll 74 is 3.0 mm, the defect period candidate probability is 94% (both sides probability of 0.2 mm / 3.0 mm = 0.07), and the probability of the outer periphery of the charging roll 74 is It is determined that it is the highest, and a defect period of 46.6 mm is output. Further, in the example of the outer periphery of the charging roll 74, the average value of the difference distance is used to calculate the probability that occurs on the outer periphery of the charging roll 74. However, after calculating the probability for each difference distance, the average value is taken. May be.

  The defect period detecting means 18 outputs the defect period having the highest defect period candidate probability, but the present invention is not limited to this, and a plurality of defect periods may be output. That is, when there is a large variation in the state of occurrence of image defects and there is a possibility of a plurality of candidates, it may be output together with probability information. For example, the outer periphery 93.2 mm of the secondary transfer roll 82 and the photosensitive drum 62 interval 95.7 mm are very close, but the detected interval may be weighted according to the distance to each and output. Good. The weighting may be simply proportional to the distance from the reference position, but is preferably set based on the distribution of detected defect periods.

  In the example of the defect interval 101.9 mm in FIG. 5A described above, the defect cycle candidate probability is 57% of the outer periphery of the photosensitive drum 62, 7% of the interval of the photosensitive drum 62, and 9% of the outer periphery of the secondary transfer roll 82. Although it may be output as it is, it is also possible to output the corrected percentage by adding the whole defect cycle candidate probabilities. In this case, in FIG. 5A, the outer periphery of the photosensitive drum 62 is 99.5 mm: 78%, the outer periphery of the secondary transfer roll 82 is 93.2 mm: 10%, and the interval between the photosensitive drums 62 is 95.7 mm: 12%. By calculating the defect cycle candidate probability in this way, it becomes possible to connect a failure diagnosis apparatus using probability information as a subsequent process of the image analysis apparatus. As will be described in detail later, when a failure diagnosis apparatus using a Bayesian network is applied, the probability information obtained by the image analysis apparatus 10 can be input to the Bayesian network so that diagnosis can be performed together with other diagnosis information. This makes it possible to improve the accuracy of fault diagnosis.

  Subsequently, a flow of processing performed by the image analysis apparatus 10 according to the first embodiment of the present invention configured as described above will be described. FIG. 6 is a flowchart showing an example of the flow of processing performed by the image analysis apparatus 10 according to the first embodiment of the present invention.

  First, in step 100, read image data obtained by reading an image based on a predetermined test chart formed on a sheet-like document by the image forming apparatus 50 by the image reading unit 54 is acquired by the image analysis apparatus 10. The process proceeds to step 102.

  In step 102, difference image data between the reference image data of the test chart and the read image data read in step 100 is calculated by the difference image data calculation means 12, and the process proceeds to step 104.

  In step 104, the projection waveform calculation means 14 calculates the projection waveform data by adding the difference image data in the main scanning direction. That is, the projection waveform data is generated by adding the image data in the main scanning direction of the difference image data. Since the projection waveform data adds data in the main scanning direction, periodic image defects in the sub-scanning direction intersecting the main scanning direction are emphasized.

  Next, in step 106, the image defect interval is detected by the defect interval detection unit 16 from the projection waveform data calculated by the projection waveform calculation unit 14, and the process proceeds to step 108. As described above, the defect interval is calculated using the position of a predetermined value with respect to the peak value (for example, the half value of the peak value) with respect to the rise of the projection waveform data of each image defect as the reference position of the image defect. The occurrence interval is detected. At this time, as described above, the moving average process may be performed on the projection waveform data.

  In step 108, defect cycle candidates are detected by the defect cycle detection unit 18 based on the image defect interval detected by the defect interval detection unit 16 and the plurality of reference values 20, and the process proceeds to step 110. The detection result detected by 18 is output to the monitor of the operation panel 56 or the like. That is, the defect period calculated by the defect interval detector 16 and the reference period of the photosensitive drum 62, the charging roll 74, the developing roll, the primary transfer roll 78, the secondary transfer roll 82, the heat roll, the intermediate transfer belt 60, and the like. The defect period candidate probability is calculated on the basis of the difference distance between them, and the period of the reference value 20 having the highest defect period candidate probability or a plurality of defect period candidate probabilities and the period of the reference value 20 is output to the operation panel 56. Is done. As a result, the output result is displayed on the monitor of the operation panel 56, the defect cycle candidate is notified to the user, and by checking this, the user performs failure diagnosis such as searching for the cause of the image defect.

(Second Embodiment)
Next, an image analysis apparatus according to the second embodiment of the present invention will be described. In the first embodiment, the defect interval detection unit 16 calculates the occurrence interval of the image defect using the position of the predetermined value with respect to the rising or falling peak value of the projection waveform data calculated by the projection waveform calculation unit 14 as an image defect. However, in the present embodiment, the self-correlation of the projection waveform data is calculated, and the occurrence interval of the image defect is obtained from the calculated autocorrelation, and the first embodiment is applied to other configurations. The difference will be described. Here, autocorrelation is used as a method of calculating the self correlation. Autocorrelation is a technique for examining the correlation between data points in series data, and the correlation coefficient is 0 means that there is no correlation between data. FIG. 7 is a diagram showing the configuration of the defect interval detection means 16 of the image analysis apparatus according to the second embodiment of the present invention.

  As shown in FIG. 7, the defect interval detection means 16 of the present embodiment includes an autocorrelation calculation unit 16A and a defect interval detection unit 16B.

  The autocorrelation calculation unit 16A calculates the autocorrelation of the projection waveform data calculated by the projection waveform calculation means 14. The autocorrelation is calculated by preparing the same projection waveform data and moving one projection waveform data by a predetermined interval (for example, one point) to calculate a correlation coefficient. Note that empty data having the same length as the projection waveform data is added after the projection waveform data in accordance with the movement of the projection waveform data.

  The autocorrelation of the projection waveform data by the autocorrelation calculation unit 16A is calculated by using the following formula for obtaining a general correlation function R.

  FIGS. 8A and 8B are diagrams showing the results of calculating autocorrelation by the autocorrelation calculation unit 16A for the projection waveform data shown in FIGS. 3A and 3B. The peak value of the correlation coefficient is a location where the projection waveform data matches, and indicates the occurrence interval of image defects.

  Therefore, the defect interval detector 16B detects the image defect interval by determining the interval between the calculated autocorrelation peak values. In the case of FIG. 8A, the defect interval is 101.8 mm. Similarly, in the case of FIG. 8B, the defect intervals are 46.8 mm, 47.4 mm, and 48.6 mm. In the first embodiment, it is desirable to perform the moving average process, but in this embodiment, it is possible to detect the image defect interval without performing the moving average process in advance, and it is possible to detect disturbances, errors, and the like. Improves stability against uncertain fluctuations.

  Next, the flow of processing performed by the image analysis apparatus according to the second embodiment of the present invention configured as described above will be described. FIG. 9 is a flowchart showing an example of the flow of processing performed by the image analysis apparatus according to the second embodiment of the present invention. Note that the same processes as those in the first embodiment will be described with the same reference numerals.

  First, in step 100, read image data obtained by reading an image based on a predetermined test chart formed on a sheet-like document by the image forming apparatus 50 by the image reading unit 54 is acquired by the image analysis apparatus, Move on to step 102.

  In step 102, difference image data between the reference image data of the test chart and the read image data read in step 100 is calculated by the difference image data calculation means 12, and the process proceeds to step 104.

  In step 104, the projection waveform calculation means 14 calculates the projection waveform data by adding the difference image data in the main scanning direction. That is, the projection waveform data is generated by adding the image data in the main scanning direction of the difference image data. Since the projection waveform data adds data in the main scanning direction, periodic image defects in the sub-scanning direction intersecting the main scanning direction are emphasized.

  Next, in step 105, the autocorrelation of the projection waveform data calculated by the projection waveform calculation means 14 is calculated by the autocorrelation calculation unit 16A, and the process proceeds to step 107. That is, as described above, the autocorrelation calculation unit 16A prepares the same projection waveform data, and calculates the autocorrelation by moving one projection waveform data by a predetermined interval and calculating the correlation coefficient.

  In step 107, the image defect interval is detected by the defect interval detector 16B by obtaining the calculated autocorrelation peak value interval, and the process proceeds to step 108.

  In step 108, defect cycle candidates are detected by the defect cycle detection unit 18 based on the image defect interval detected by the defect interval detection unit 16 and the plurality of reference values 20, and the process proceeds to step 110. The detection result detected by 18 is output to the monitor of the operation panel 56 or the like. That is, the defect period calculated by the defect interval detector 16 and the reference period of the photosensitive drum 62, the charging roll 74, the developing roll, the primary transfer roll 78, the secondary transfer roll 82, the heat roll, the intermediate transfer belt 60, and the like. The defect period candidate probability is calculated on the basis of the difference distance between them, and the period of the reference value 20 having the highest defect period candidate probability or a plurality of defect period candidate probabilities and the period of the reference value 20 is output to the operation panel 56. Is done. As a result, the output result is displayed on the monitor of the operation panel 56, the defect cycle candidate is notified to the user, and by checking this, the user performs failure diagnosis such as searching for the cause of the image defect.

(Third embodiment)
Next, an image analysis apparatus according to a third embodiment of the present invention will be described. In this embodiment, in order to remove noise and accurately detect the interval between image defects, the difference image data is binarized in advance and the projection waveform data is calculated. The first embodiment or the second embodiment can be applied. In addition, although 2nd Embodiment is applied in this embodiment, you may make it apply 1st Embodiment.

  FIG. 10 is a block diagram showing a detailed configuration of the projection waveform calculation means 14 according to the third embodiment of the present invention.

  As shown in FIG. 10, the projection waveform calculation means 14 of the present embodiment includes a binarization processing unit 14C that binarizes the difference image data before performing the projection waveform calculation process. After the difference image data is binarized by 14C, the projection waveform calculation processing unit 14A calculates the projection waveform data of the difference image data in the main scanning direction. The projection waveform calculation processing unit 14A calculates projection waveform data in the same manner as the projection waveform calculation unit 14 described in the first embodiment. Further, in the present embodiment, a description will be made assuming that the binarization processing unit 14C is included in the projection waveform calculation unit 14, but the difference image data calculation unit 12 and the projection waveform calculation unit 14 in the first embodiment and the second embodiment are between. A binarization processing unit 14C may be provided.

  FIG. 11A is a diagram showing the projection waveform data when the projection waveform data is calculated after the difference image data is binarized with respect to the projection waveform data of FIG. 3B. As can be seen from FIG. 11A, the binarization process reduces the influence of background noise on a clear defect, and the image defect interval becomes clear. FIG. 11B is a diagram showing a result of calculating the autocorrelation corresponding to the projection waveform data of FIG. 11A, and the peak value of the correlation coefficient is larger than that of FIG. 8B. It becomes steep and it can be seen that the influence of the background noise is reduced. In the autocorrelation calculation result of FIG. 11B, the defect interval is 46.5 mm, 47.1 mm, 46.3 mm, and an average of 46.6 mm, which coincides with the outer periphery of the charging roll 74 of 46.6 mm, and is detected with higher accuracy. I understand that.

  Next, the flow of processing performed by the image analysis apparatus according to the third embodiment of the present invention configured as described above will be described. FIG. 12 is a flowchart showing an example of the flow of processing performed by the image analysis apparatus according to the third embodiment of the present invention. Note that the same processes as those in the second embodiment are denoted by the same reference numerals.

  First, in step 100, read image data obtained by reading an image based on a predetermined test chart formed on a sheet-like document by the image forming apparatus 50 by the image reading unit 54 is acquired by the image analysis apparatus, Move on to step 102.

  In step 102, difference image data between the reference image data of the test chart and the read image data read in step 100 is calculated by the difference image data calculation unit 12, and the process proceeds to step 103.

  In step 103, the difference image data is binarized by the binarization processing unit 14C, and the process proceeds to step 104.

  In step 104, the projection waveform calculation means 14 calculates the projection waveform data by adding the difference image data in the main scanning direction. That is, in this embodiment, the projection waveform calculation processing unit 14A adds the image data in the main scanning direction of the difference image data binarized by the binarization processing unit 14C, thereby generating projection waveform data. The Since the projection waveform data adds data in the main scanning direction, periodic image defects in the sub-scanning direction intersecting the main scanning direction are emphasized.

  Next, in step 105, the autocorrelation of the projection waveform data calculated by the projection waveform calculation means 14 is calculated by the autocorrelation calculation unit 16A, and the process proceeds to step 107. That is, as described above, the autocorrelation calculation unit 16A prepares the same projection waveform data, and calculates the autocorrelation by moving one projection waveform data by a predetermined interval and calculating the correlation coefficient.

  In step 107, the image defect interval is detected by the defect interval detector 16B by obtaining the calculated autocorrelation peak value interval, and the process proceeds to step 108.

  In step 108, defect cycle candidates are detected by the defect cycle detection unit 18 based on the image defect interval detected by the defect interval detection unit 16 and the plurality of reference values 20, and the process proceeds to step 110. The detection result detected by 18 is output to the monitor of the operation panel 56 or the like. That is, the defect period calculated by the defect interval detector 16 and the reference period of the photosensitive drum 62, the charging roll 74, the developing roll, the primary transfer roll 78, the secondary transfer roll 82, the heat roll, the intermediate transfer belt 60, and the like. The defect period candidate probability is calculated on the basis of the difference distance between them, and the period of the reference value 20 having the highest defect period candidate probability or a plurality of defect period candidate probabilities and the period of the reference value 20 is output to the operation panel 56. Is done. As a result, the output result is displayed on the monitor of the operation panel 56, the defect cycle candidate is notified to the user, and by checking this, the user performs failure diagnosis such as searching for the cause of the image defect.

  Such an image analysis apparatus according to the embodiment of the present invention has, for example, an image forming apparatus in which there is no clear boundary of the generation cycle as compared with an image defect caused by a flaw of a photosensitive drum or a fixing device roll, and the image The present invention may be applied to periodic image defects such as charging defects due to wear or dirt on image forming members having large unevenness within the defects, streaky dirt or belt-like dirt generated due to transfer defects or cleaning defects.

  Note that the image analysis device according to each of the above embodiments may be incorporated in a failure diagnosis device using a Bayesian network. Below, the outline in the case of incorporating an image analysis apparatus in a failure diagnosis apparatus using a Bayesian network will be described. As a detailed configuration of the failure diagnosis apparatus described below, for example, the technique described in Japanese Patent Application Laid-Open No. 2005-309077 or Japanese Patent Application Laid-Open No. 2005-309078 proposed by the applicant of the present application is applied. In the following, the latter case will be described as an example.

  FIG. 13 is a block diagram showing a configuration example when the image analysis apparatus according to the embodiment of the present invention is incorporated in a failure diagnosis apparatus using a Bayesian network.

  The failure diagnosis apparatus 200 shown in FIG. 13 is configured to perform failure diagnosis by acquiring various information from the image forming apparatus 50.

  For example, as shown in FIG. 13, an image analysis apparatus 222 according to the embodiment of the present invention, a component state information acquisition unit 202 that acquires component information indicating the operation status of components as observation data information, and an image forming apparatus 50. The history information acquisition management unit 204 that manages history information by monitoring and monitoring the usage status and registering and holding the monitoring results in a nonvolatile storage medium, and the temperature and humidity based on the temperature and humidity information during operation Based on the environmental information acquisition unit 206 that acquires environmental information that affects the state of components such as environmental information as environmental information and the detection information of consumables, the thickness and type of printing paper, or the color type and type of coloring material, A consumable material information acquisition unit 208 that acquires information on consumable materials used by the apparatus such as the remaining amount, and a specification information acquisition unit 2 that acquires specification information of the image forming apparatus 50 And a 0, a.

  In addition, the failure diagnosis apparatus 200 obtains a diagnosis model based on a feature quantity extraction unit 212 that acquires various types of information from each of the acquisition means described above and extracts a feature quantity, and an image defect extraction unit that is detected by the image analysis apparatus 222. A Bayesian network diagnostic model database 216 for selection, a Bayesian network inference processing unit 218 that identifies a failure location based on the selected diagnostic model and various pieces of feature amount information input as evidence information from the feature amount extraction unit, A reference feature amount storage unit 214 that stores a reference feature amount serving as a determination index at the time of diagnosis in a predetermined storage medium (preferably a non-volatile semiconductor memory), and a notification unit 220 that notifies a user of a failure determination result and inspection contents It is equipped with.

  The feature amount extraction unit 212 selects a failure diagnosis model related to the image defect based on the image defect feature amount output from the image analysis device 222, and performs diagnosis necessary for the image defect diagnosis from the Bayesian network diagnosis model database 216. Extract information items. Further, the feature quantity extraction unit 212 acquires the component state information acquisition unit 202, the history information acquisition management unit 204, the environment information acquisition unit 206, the consumable material information acquisition unit 208, or the specification information acquisition based on the extracted diagnostic information item. The diagnostic feature quantity information necessary for the image defect diagnosis is acquired from the unit 210. Furthermore, for the predetermined diagnostic feature amount information input as an analog value in the acquired diagnostic feature amount information, the diagnosis is performed based on threshold information related to the feature amount output from the reference feature amount storage unit 214. Discretize information.

  Here, the Bayesian network used in the Bayesian network inference processing unit 218 will be briefly described. FIG. 14 is a Bayesian network model diagram illustrating a basic configuration example of a Bayesian network.

  A Bayesian network is a directed acyclic graph representing a causal relationship between variables, and associates a conditional probability distribution with a variable when a parent is given. Bayesian networks model problem areas using probability theory. Given information about others, a Bayesian network representation of the problem is used to provide information about a subset of variables.

  A Bayesian network is composed of a set of variables (nodes: indicated by ellipses in FIG. 14) and a set of arcs (indicated by arrows in FIG. 14) indicating directed edges (connections between variables). Arrows called arcs indicate causal relationships and are linked from cause to result in the direction of the arrow.

  Each node (variable) has a set of mutually exclusive states. For each no, a probability (conditional probability table) that a result occurs from a cause is set in advance. The nodes form a directed acyclic graph (DAG) with a directed edge. A conditional probability table P (v | w1,..., Wn) is defined for each variable v having parents w1,.

  In FIG. 14, the nodes indicated by hatching are nodes that can be directly observed. By calculating the probability of a node shown without hatching, the state of the component (possibility of failure) can be determined. For example, Bayes' theorem is used for the probability calculation of each node. However, in a network configuration in which there are many nodes and a loop is formed, the calculation is practically impossible due to the enormous amount of calculation. Therefore, various efficient calculation algorithms for accurately updating the probability in the Bayesian network have been devised, and calculation software is also sold by several manufacturers.

  When performing failure determination with such a failure diagnosis device 200, as shown in FIG. 13, when the failure probability is inferred by the Bayesian network inference processing unit 218, an image analysis device to which any of the above-described embodiments is applied. By inputting the defect period candidate probabilities obtained at 222 to the Bayesian network inference processing unit 218, it becomes possible to perform failure diagnosis together with other diagnosis information as described above, and failure diagnosis with higher accuracy than the prior art. Is possible.

  Electrophotographic copiers and printers should operate in extremely harsh machine environments such as charging / development / transfer with high-voltage power supplies, high-temperature fixing, polymer toner scattering, high-speed paper transport, and paper dust. In order to maintain good quality, it is necessary to perform regular maintenance. The user of an electrophotographic copying machine or printer diagnoses the failure part. There is a growing demand to reduce service costs by exchanging and repairing parts, or by contacting service personnel with accurate failure information. By using the above-described failure diagnosis device using the Bayesian network, various types of information necessary for failure diagnosis can be acquired accurately and without stressing users who do not have specialized knowledge. Diagnosis is possible.

  Note that the functions of the failure diagnosis apparatus 200 using the image analysis apparatus or the Bayesian network according to each of the above embodiments may be realized by a computer program. FIG. 15 shows an example of a computer program, a storage medium storing the computer program, and an example of a computer when the functions of the image analysis apparatus according to the embodiment of the present invention and the failure diagnosis apparatus 200 using a Bayesian network are realized by a computer program. It is explanatory drawing of. In the figure, 150 is a program, 152 is a computer, 154 is a magneto-optical disk, 156 is an optical disk, 158 is a magnetic disk, 160 is a memory, 162 is an internal memory, 166 is a reading unit, 170 is a hard disk, 168 and 174 are interfaces, Reference numeral 172 denotes a communication unit.

  Part or all of the functions of each unit of the failure diagnosis apparatus 200 using the image analysis apparatus or the Bayesian network according to the embodiment of the present invention described above can be realized by a program 150 that can be executed by a computer. In that case, the program 150 and data used by the program can be stored in a computer-readable storage medium. As a storage medium, the reading unit 166 provided in the hardware resource of the computer causes a change state of energy such as magnetism, light, electricity, etc. according to the description content of the program, and a signal corresponding thereto is generated. In the format, the description content of the program can be transmitted to the reading unit 166. For example, a magneto-optical disk 154, an optical disk 156 (including a CD and a DVD), a magnetic disk 158, a memory 160 (including an IC card and a memory card), and the like. Of course, these storage media are not limited to portable types.

  The program 150 is stored in these storage media, and the program 150 is read from the computer and stored in the internal memory 162 or the hard disk 170, for example, by mounting these storage media on the reading unit 166 or the interface 174 of the computer 152. By executing the program 150 by the CPU 164, the function of the failure diagnosis apparatus 200 using the image analysis apparatus or the Bayesian network according to the embodiment of the present invention can be realized. Alternatively, the program 150 is transferred to the computer 152 via a network or the like, and the computer 152 receives the program 150 by the communication unit 172 and stores it in the internal memory 162 or the hard disk 170, and the program 150 is executed by the CPU 164. You may implement | achieve the function of the failure diagnosis apparatus 200 using the image-analysis apparatus which concerns on embodiment of an invention, or a Bayesian network. In addition, the computer 152 can be connected to various devices via an interface 168. For example, a display device for displaying information and an input device for inputting information by a user are also connected.

  Of course, some functions may be configured by hardware, or all may be configured by hardware. Alternatively, it can be configured as a program including the embodiment of the present invention together with other configurations.

  In the above embodiment, the case of detecting the period of periodic image defects in the sub-scanning direction that occurs when an image is formed by the image forming apparatus 50 has been described as an example. However, the present invention is not limited thereto. The embodiment of the present invention can be applied as long as it detects the period of a periodic image even if it is not an image defect.

  Further, in the above-described embodiment, it has been described that the period of an image defect having a periodicity in the sub-scanning direction of an image formed by an electrophotographic image forming apparatus is detected. The period of a periodic image in an image formed by an image forming apparatus using LPH (LED Print Head) may be detected. In this case, when a driver for driving the LEDs is provided for a predetermined number of LEDs, an image having periodicity in the main scanning direction may be formed. Therefore, a periodic image generated in the main scanning direction. This period can also be detected. That is, the periodic direction of the periodic image is not particularly limited to the above-described embodiment (image defect having periodicity in the sub-scanning direction).

1 is a diagram showing a schematic configuration of an image forming apparatus for recording an image for detecting an image defect by an image analysis apparatus according to an embodiment of the present invention. 1 is a block diagram illustrating a schematic configuration of an image analysis apparatus according to a first embodiment of the present invention. (A) is a figure which shows an example of the projection waveform of the image defect of the outer periphery period of a photoconductor drum, (B) is a figure which shows an example of the projection waveform of the image defect of the outer periphery period of a charging roll. It is a block diagram which shows a structure at the time of adding a moving average process to a projection waveform calculation means. It is a figure which shows an example of the projection waveform data at the time of performing the averaging process for every 75 points with respect to FIG. 3 (A) and (B). It is a flowchart which shows an example of the flow of the process performed with the image analysis apparatus which concerns on 1st Embodiment of this invention. It is a figure which shows the structure of the defect space | interval detection means of the image analyzer which concerns on 2nd Embodiment of this invention. It is a figure which shows the result of having calculated the autocorrelation by the autocorrelation calculation part with respect to the projection waveform data shown to FIG. 3 (A), (B). It is a flowchart which shows an example of the flow of the process performed with the image analysis apparatus which concerns on 2nd Embodiment of this invention. It is a block diagram which shows the detailed structure of the projection waveform calculation means which concerns on 3rd Embodiment of this invention. It is a figure which shows the projection waveform data at the time of calculating projection waveform data, after carrying out the binarization process of the difference image data with respect to the projection waveform data of FIG. 3 (B). It is a flowchart which shows an example of the flow of the process performed with the image analysis apparatus which concerns on 3rd Embodiment of this invention. It is a block diagram which shows the structural example at the time of incorporating the image-analysis apparatus which concerns on embodiment of this invention in the failure diagnosis apparatus using a Bayesian network. It is a Bayesian network model figure which shows the basic structural example of a Bayesian network. FIG. 3 is an explanatory diagram of an example of a computer program, a storage medium storing the computer program, and an example of a computer when the functions of the failure diagnosis device using the image analysis device and the Bayesian network according to the embodiment of the present invention are realized by the computer program . (A) is a figure which shows an example of a blank test chart, (B) is a figure which shows an example of the test chart containing the image defect with the periodicity with respect to the blank test chart. It is a figure which shows the difference image of FIG. 16 (A) and FIG. 16 (B).

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Image analyzer 12 Difference image data calculation means 14 Projected waveform calculation means 14A Projected waveform calculation part 14B Moving average process part 14C Binarization process part 16 Defect interval detection means 16A Autocorrelation calculation part 16B Defect interval detection part 18 Defect period detection Means 20 Reference value 50 Image forming apparatus 56 Operation panel 150 Program 152 Computer 200 Fault diagnosis apparatus

Claims (19)

A projection waveform calculating means for calculating a projection waveform in a direction intersecting with a periodic direction of periodicity of image data having image characteristics including at least periodicity;
A periodic interval detecting means for detecting a periodic interval of periodicity included in the image data based on a predetermined edge position of the projected waveform calculated by the projected waveform calculating means;
An image analysis apparatus comprising: A projection waveform calculating means for calculating a projection waveform in a direction intersecting with a periodic direction of periodicity of image data having image characteristics including at least periodicity;
A period interval detection unit for obtaining a self correlation of the projection waveform calculated by the projection waveform calculation unit, and detecting a periodic interval of periodicity included in the image data based on the correlation;
An image analysis apparatus comprising: Difference calculating means for calculating difference image data between image data having image characteristics including at least periodicity and predetermined image data;
A projection waveform calculation unit that calculates a projection waveform in a direction that intersects the periodic direction of the periodicity of the image data from the difference image data calculated by the difference calculation unit;
A periodic interval detecting means for detecting a periodic interval of periodicity included in the image data based on a predetermined edge position of the projected waveform calculated by the projected waveform calculating means;
An image analysis apparatus comprising: Difference calculating means for calculating difference image data between image data having image characteristics including at least periodicity and predetermined image data;
A projection waveform calculation unit that calculates a projection waveform in a direction that intersects the periodic direction of the periodicity of the image data from the difference image data calculated by the difference calculation unit;
A period interval detection unit for obtaining a self correlation of the projection waveform calculated by the projection waveform calculation unit, and detecting a periodic interval of periodicity included in the image data based on the correlation;
An image analysis apparatus comprising:   Based on the periodic interval detected by the periodic interval detecting means and a plurality of predetermined reference values representing the period of a component that causes a periodic image in the image forming apparatus that forms an image, the image data 2. A detection means for calculating a candidate probability for each reference value of periodicity, which is an image feature included, and detecting a periodicity period included in the image data based on the calculated candidate probability. The image analysis device according to claim 1.   6. The projection waveform calculation unit according to claim 1, wherein the projection waveform calculation unit calculates a projection waveform by performing an addition process on a direction intersecting with a periodic direction of the image data. The image analysis apparatus described in 1.   The image analysis apparatus according to claim 1, wherein the projection waveform calculation unit includes a moving average process for calculating a moving average of the calculated projection waveform.   8. The binarization unit according to claim 1, further comprising: binarization unit that binarizes the image data when the projection waveform calculation unit calculates the projection waveform. 9. Image analysis device.   The detection means detects the reference value corresponding to the highest candidate probability among the calculated candidate probabilities as a periodicity period included in the image data. The image analysis apparatus according to any one of the above.   9. The detection unit according to claim 5, wherein the detection unit detects a plurality of reference values corresponding to the calculated candidate probabilities as periodic periods included in the image data. 10. The image analysis apparatus described in 1.   The image analysis apparatus according to claim 1, further comprising a notification unit that notifies a detection result of the detection unit. Difference calculating means for calculating difference image data between image data having image characteristics including at least periodicity and predetermined image data;
A projection waveform calculation means for calculating a projection waveform in the periodic direction of the periodicity of the image data from the difference image data calculated by the difference calculation means;
A periodic interval detecting means for detecting a periodic interval of periodicity included in the image data based on a predetermined edge position of the projected waveform calculated by the projected waveform calculating means;
Based on the periodic interval detected by the periodic interval detecting means and a plurality of predetermined reference values representing the period of a component that causes a periodic image in the image forming apparatus that forms an image, the image data Candidate probability calculating means for calculating a candidate probability for each reference value of periodicity that is an image feature included;
A failure diagnosis means for performing a failure diagnosis of the image forming apparatus by inputting the candidate probability calculated by the candidate probability calculation means into a probability model;
Fault diagnosis device with Difference calculating means for calculating difference image data between image data having image characteristics including at least periodicity and predetermined image data;
A projection waveform calculation unit that calculates a projection waveform in a direction that intersects the periodic direction of the periodicity of the image data from the difference image data calculated by the difference calculation unit;
A period interval detection unit for obtaining a self correlation of the projection waveform calculated by the projection waveform calculation unit, and detecting a periodic interval of periodicity included in the image data based on the correlation;
Based on the periodic interval detected by the periodic interval detecting means and a plurality of predetermined reference values representing the period of a component that causes a periodic image in the image forming apparatus that forms an image, the image data Candidate probability calculating means for calculating a candidate probability for each reference value of periodicity that is an image feature included;
A failure diagnosis means for performing a failure diagnosis of the image forming apparatus by inputting the candidate probability calculated by the candidate probability calculation means into a probability model;
Fault diagnosis device with A projection waveform calculating step for calculating a projection waveform in a direction intersecting with the periodic direction of the periodicity of the image data having image characteristics including at least periodicity;
A periodic interval detecting step of detecting a periodic interval of periodicity included in the image data based on a predetermined edge position of the projected waveform calculated by the projected waveform calculating step;
An image analysis program for causing a computer to execute a process including: A projection waveform calculating step for calculating a projection waveform in a direction intersecting with the periodic direction of the periodicity of the image data having image characteristics including at least periodicity;
A self-correlation of the projection waveform calculated by the projection waveform calculation step, and a periodic interval detection step of detecting a periodic interval of periodicity included in the image data based on the correlation;
An image analysis program for causing a computer to execute a process including: A difference calculating step for calculating difference image data between image data having image characteristics including at least periodicity and predetermined image data;
A projection waveform calculation step of calculating a projection waveform in a direction intersecting with the periodic direction of the periodicity of the image data from the difference image data calculated by the difference calculation step;
A periodic interval detecting step for detecting a periodic interval of periodicity included in the image data based on a predetermined edge position of the projected waveform calculated by the projected waveform calculating step;
An image analysis program for causing a computer to execute a process including: A difference calculating step for calculating difference image data between image data having image characteristics including at least periodicity and predetermined image data;
A projection waveform calculation step of calculating a projection waveform in a direction intersecting with the periodic direction of the periodicity of the image data from the difference image data calculated by the difference calculation step;
A self-correlation of the projection waveform calculated by the projection waveform calculation step, and a periodic interval detection step of detecting a periodic interval of periodicity included in the image data based on the correlation;
An image analysis program for causing a computer to execute a process including: A difference calculating step for calculating difference image data between image data having image characteristics including at least periodicity and predetermined image data;
A projection waveform calculation step of calculating a projection waveform in a direction intersecting with the periodic direction of the periodicity of the image data from the difference image data calculated by the difference calculation step;
A periodic interval detecting step for detecting a periodic interval of periodicity included in the image data based on a predetermined edge position of the projected waveform calculated by the projected waveform calculating step;
Based on the periodic interval detected by the periodic interval detecting step and a plurality of predetermined reference values representing the period of a component that causes a periodic image in the image forming apparatus that forms an image, the image data A candidate probability calculating step of calculating a candidate probability for each reference value of periodicity that is an image feature included;
A fault diagnosis step of performing a fault diagnosis of the image forming apparatus by inputting the candidate probability calculated by the candidate probability calculation step into a probability model;
A failure diagnosis program for causing a computer to execute a process including: A difference calculating step for calculating difference image data between image data having image characteristics including at least periodicity and predetermined image data;
A projection waveform calculation step of calculating a projection waveform in a direction intersecting with the periodic direction of the periodicity of the image data from the difference image data calculated by the difference calculation step;
A self-correlation of the projection waveform calculated by the projection waveform calculation step, and a periodic interval detection step of detecting a periodic interval of periodicity included in the image data based on the correlation;
Based on the periodic interval detected by the periodic interval detecting step and a plurality of predetermined reference values representing the period of a component that causes a periodic image in the image forming apparatus that forms an image, the image data A candidate probability calculating step of calculating a candidate probability for each reference value of periodicity that is an image feature included;
A fault diagnosis step of performing a fault diagnosis of the image forming apparatus by inputting the candidate probability calculated by the candidate probability calculation step into a probability model;
A failure diagnosis program for causing a computer to execute a process including: