JP2022102540A - Belt inspection system and belt inspection program - Google Patents

Abstract

To provide a belt inspection system and a belt inspection program that can improve accuracy of detecting a belt defect.SOLUTION: A belt inspection system comprises a defect candidate detection unit that detects, from a belt image as an image of an intermediate transfer belt of an image forming apparatus, candidates for a belt defect as an abnormal part of the intermediate transfer belt. The defect candidate detection unit executes a pre-processing step of executing, on the belt image, pre-processing for detecting the candidate for the belt defect, and a defect candidate detection step of detecting the candidate for the belt defect from the belt image on which the pre-processing is executed. In the pre-processing step, the defect candidate detection unit divides brightness values of respective pixels of the belt image by the brightness values of corresponding pixels in an image obtained by removing a high frequency component from the belt image to remove in-plane unevenness from the belt image (S163).SELECTED DRAWING: Figure 9

Description

The present invention relates to a belt inspection system and a belt inspection program for detecting an abnormal portion (hereinafter referred to as "belt defect") of an intermediate transfer belt of an image forming apparatus.

As a conventional belt inspection system, a system is known in which the surface of an intermediate transfer belt is optically read by an image pickup device and belt defects are automatically detected based on the reading result (see, for example, Patent Document 1).

Japanese Unexamined Patent Publication No. 2012-014108

However, in the conventional belt inspection system, there is a problem that the belt defect cannot be detected with high accuracy.

Therefore, an object of the present invention is to provide a belt inspection system and a belt inspection program capable of improving the accuracy of detecting belt defects.

The belt inspection system of the present invention includes a defect candidate detection unit that detects a candidate for a belt defect as an abnormal portion of the intermediate transfer belt from a belt image as an image of the intermediate transfer belt of the image forming apparatus, and the defect candidate detection unit. Executes a pretreatment step of applying a pretreatment for detecting the candidate to the belt image and a defect candidate detection step of detecting the candidate from the belt image to which the pretreatment has been performed. In the preprocessing step, the detection unit divides the brightness value of each pixel of the belt image by the brightness value of the corresponding pixel in the image obtained by removing the high frequency component from the belt image, thereby causing in-plane unevenness from the belt image. It is characterized by removing.

With this configuration, the belt inspection system of the present invention removes the brightness value of each pixel of the belt image and the high frequency component from the belt image in the pretreatment step of applying the pretreatment for detecting the candidate of the belt defect to the belt image. Since in-plane unevenness is removed from the belt image by dividing by the brightness value of the corresponding pixel in the image, the brightness of the corresponding pixel in the image obtained by removing the high frequency component from the belt image from the brightness value of each pixel in the belt image. By subtracting the value, the in-plane unevenness can be appropriately removed from the belt image as compared with the case where the in-plane unevenness is removed from the belt image, and as a result, the accuracy of detecting the belt defect can be improved. can.

In the belt inspection system of the present invention, the defect candidate detection unit is based on the pixels of the belt image having a brightness lower than the average of the brightness of the belt image and not more than a specific brightness in the defect candidate detection step. The candidate may be detected.

With this configuration, the belt inspection system of the present invention detects candidates for belt defects based on the pixels of the belt image whose brightness is below a specific brightness below the average of the brightness of the belt image. Belt defects with lightness below the average lightness can be detected.

In the belt inspection system of the present invention, the defect candidate detection unit obtains a value obtained by subtracting the average value of the brightness values of the belt image from the brightness value of each pixel of the belt image in the defect candidate detection step. A standardized value is calculated by dividing by the standard deviation of the brightness value of the image, and the pixels having a brightness equal to or lower than the specific brightness among the pixels of the belt image are the standardized values among the pixels of the belt image. It may be a pixel whose value is equal to or less than a specific negative value.

With this configuration, the belt inspection system of the present invention detects candidates for belt defects based on the pixels whose standardized brightness value is equal to or less than a specific negative value among the pixels of the belt image. The influence of individual differences in the intermediate transfer belt on the detection accuracy can be reduced.

The belt inspection program of the present invention realizes a defect candidate detection unit for detecting a candidate for a belt defect as an abnormal portion of the intermediate transfer belt from a belt image as an image of the intermediate transfer belt of the image forming apparatus on a computer, and realizes the defect. The candidate detection unit executes a pretreatment step of applying a pretreatment for detecting the candidate to the belt image and a defect candidate detection step of detecting the candidate from the belt image to which the pretreatment has been performed. In the preprocessing step, the defect candidate detection unit divides the brightness value of each pixel of the belt image by the brightness value of the corresponding pixel in the image from which the high frequency component is removed from the belt image, so that the defect candidate detection unit can use the belt image. It is characterized by removing in-plane unevenness.

With this configuration, the computer executing the belt inspection program of the present invention obtains the brightness value of each pixel of the belt image from the belt image in the preprocessing step of applying the preprocessing for detecting the candidate of the belt defect to the belt image. In-plane unevenness is removed from the belt image by dividing by the brightness value of the corresponding pixel in the image from which the high frequency component has been removed. By subtracting the brightness value of the pixel to be used, the in-plane unevenness can be appropriately removed from the belt image as compared with the case where the in-plane unevenness is removed from the belt image, and as a result, the accuracy of detecting the belt defect can be improved. Can be improved.

The belt inspection system and the belt inspection program of the present invention can improve the accuracy of detecting belt defects.

It is a figure which shows the example of the needle-like "coating defect" which is a belt defect of the intermediate transfer belt which concerns on one Embodiment of this invention. It is a figure which shows the example of the fibrous "coating defect" which is a belt defect of the intermediate transfer belt which concerns on one Embodiment of this invention. It is a flowchart of the belt inspection process which concerns on one Embodiment of this invention. It is a figure which shows an example of the image pickup system which takes a picture of an intermediate transfer belt in the picture-taking process shown in FIG. It is a figure which shows an example which is different from the example shown in FIG. 4 of the image pickup apparatus shown in FIG. It is a block diagram of the belt inspection system which executes the diagnostic imaging process shown in FIG. 3 when it is realized by one computer. It is a flowchart of the image diagnosis process shown in FIG. It is a flowchart of the pretreatment process shown in FIG. 7. It is a figure which shows an example of the in-plane unevenness removal process shown in FIG. (A) For the purpose of explaining the in-plane unevenness removing step shown in FIG. 9, it is a schematic diagram which shows an example of the brightness value in a specific line of a belt image. (B) It is a schematic diagram which shows an example of the brightness value when the high frequency component is removed with respect to the brightness value shown in FIG. 10 (a). (A) For the purpose of explaining the in-plane unevenness removing step shown in FIG. 9, it is a figure which shows the signal which superposed the low frequency component and the high frequency component which increases and decreases in correlation with the increase and decrease of a low frequency component. (B) It is a figure which shows the signal obtained by subtracting a low frequency component from the signal shown in FIG. 11 (a). (C) It is a figure which shows the signal obtained by dividing the signal shown in FIG. 11A by a low frequency component. It is a schematic diagram which shows an example of the brightness value when the in-plane unevenness is corrected with respect to the brightness value shown in FIG. 10 (a). It is a flowchart of an example of "other pretreatment process" shown in FIG. It is a figure which shows an example of the defect candidate detection process shown in FIG. 7. It is a figure which shows an example of the standardization process shown in FIG. It is a figure which shows the probability density of a normal distribution used in the defect candidate detection process shown in FIG. It is a flowchart of the quality determination process shown in FIG. 7. For the purpose of explaining the quality determination process shown in FIG. 7, the difference between the minimum value of the lightness value in the defect candidate image and the average value of the lightness value in the preprocessed belt image obtained from the defect candidate image, and this preprocessing. It is a figure which shows an example of the relationship with the standard deviation of a lightness value in a finished belt image. FIG. 7 is a flowchart of a reference use defect determination method for determining the type and degree of defect candidates for coating defects in the quality determination step shown in FIG. 7. It is a flowchart of the standard use quality determination method in the quality determination process shown in FIG. 7.

Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

First, the intermediate transfer belt according to the present embodiment will be described.

In a color image forming apparatus, a component called an intermediate transfer belt to which a charged toner is attached is used. For example, when the image forming apparatus is an A3 machine, the intermediate transfer belt has a cylindrical shape having a width slightly larger than 297 mm, which is the length in the lateral direction of an A3 size recording medium, for example, about 330 mm, and has an outer circumference. The length of the A3 size recording medium is slightly larger than twice the length of 420 mm in the longitudinal direction, for example, about 850 mm.

The intermediate transfer belt is manufactured by inflation molding, for example, from the viewpoint of manufacturing cost, although there are several manufacturing methods. Also, from the viewpoint of manufacturing cost, the intermediate transfer belt is often manufactured by dispersing carbon in a resin and adjusting the electric resistance in the semiconductor region. When the intermediate transfer belt is manufactured by dispersing carbon in a resin, it often has a dark appearance such as black or dark brown.

The intermediate transfer belt is an electronically very delicate component that repeatedly adheres to and peels off charged toner, and is particularly severe against surface protrusions, scratches, and deformation. For example, even if the surface of the intermediate transfer belt has a deformation having a diameter of several mm and a height of about 10 μm, it affects the adhesion state of the toner, and as a result, the recording medium is used by the image forming apparatus. It deteriorates the quality of the image printed on.

Therefore, it is desired to be able to detect the belt defect as an abnormal part of the intermediate transfer belt with high accuracy.

For example, as a belt defect, there is a belt defect (hereinafter referred to as “coating defect”) that is attached to the surface of the intermediate transfer belt during spray coating on the surface of the intermediate transfer belt.

Hereinafter, the "defects of coating material" will be described in detail.

The surface of the intermediate transfer belt is coated with a coating liquid such as PTFE (poly-tetra-fluoro-ethylene) by a spray gun. When the surface of the intermediate transfer belt is coated with the coating liquid, if the viscosity of the coating liquid changes or becomes non-uniform due to the temperature and humidity conditions, etc., and the adjustment of the nozzle of the spray gun becomes incompatible, it is complete. The coating liquid that has become atomized and cannot be completely collected accumulates slightly at the tip of the nozzle. Then, the lump of the coating liquid collected at the tip of the nozzle may fly from the tip of the nozzle at some beat, and when it flies from the tip of the nozzle, it may adhere to the intermediate transfer belt and solidify.

If the size of the lump of the coating liquid adhering to the intermediate transfer belt is small, the effect on the charged state of the intermediate transfer belt is often negligible, but if the size exceeds a certain level, the effect that cannot be ignored on the charged state of the intermediate transfer belt is negligible. To exert. Further, even if the size of the coating liquid adhered to the intermediate transfer belt is small, for example, the diameter is about 100 μm, it flies in a needle-like semi-solidified state and pierces the surface of the intermediate transfer belt. It may adhere and solidify almost upright, and in that case, even if the size is small, the charge tends to concentrate, which may have a non-negligible effect on the charged state of the intermediate transfer belt.

1 (a) to 1 (e) are views showing an example of a needle-shaped "coating defect" which is a belt defect of the intermediate transfer belt according to the embodiment of the present invention. 2 (a) to 2 (e) are views showing an example of a fibrous "coating defect" which is a belt defect of the intermediate transfer belt according to the embodiment of the present invention.

Some of the "defects of coating" have a diameter of about several mm, many have a diameter of about several hundred μm, and some have a narrowest portion of 100 μm or less. The "painting defects" include, for example, needle-like ones as shown in FIGS. 1 (a) and 1 (e) and fibrous ones as shown in FIGS. 2 (a) to 2 (e). ..

Next, a belt inspection process for inspecting the intermediate transfer belt will be described.

FIG. 3 is a flowchart of the belt inspection process according to the present embodiment.

As shown in FIG. 3, the abnormality diagnosis by the image of the intermediate transfer belt is performed by the "imaging step" (S101) for photographing the intermediate transfer belt and the image of the intermediate transfer belt obtained by the imaging step of S101 (hereinafter, "" It is composed of an "imaging diagnosis step" (S102) for diagnosing an abnormality of the intermediate transfer belt based on "belt image").

FIG. 4 is a diagram showing an example of an imaging system 20 that photographs the intermediate transfer belt 10 in the imaging process of S101.

As shown in FIG. 4, the image pickup system 20 includes an image pickup device 21 for photographing the intermediate transfer belt 10 and a belt moving device 22 for moving the intermediate transfer belt 10 with respect to the image pickup device 21. The image pickup apparatus 21 is preferably installed parallel to the width direction indicated by the arrow 10a of the intermediate transfer belt 10. The image pickup device 21 extends in the extension direction of the image pickup device 21 and extends in the extension direction of the image pickup device 21 and a light source (not shown) for irradiating the surface of the intermediate transfer belt 10 with light. It is provided with an image pickup unit (not shown) such as a line sensor for photographing the surface of the intermediate transfer belt 10. The image pickup apparatus 21 is preferably installed as close as possible to the surface of the intermediate transfer belt 10.

The angle of view size of the image pickup apparatus 21 is usually set to be larger than the width of the intermediate transfer belt 10. Therefore, at the end of the intermediate transfer belt 10, a non-belt portion that is not a portion of the intermediate transfer belt 10 is also reflected. However, since the non-belt portion is not a target for inspection of belt defects, it is excluded from the processing target by being cut off or masked.

The captured image of the intermediate transfer belt 10 in the imaging step of S101 is generated by synthesizing the images captured line by line by the image pickup apparatus 21. In the image taken by the intermediate transfer belt 10 in the photographing step of S101, the length in the direction corresponding to the circumferential direction of the intermediate transfer belt 10 is longer than the length of one round of the intermediate transfer belt 10. That is, the captured image has overlapping portions at both ends in the direction corresponding to the circumferential direction of the intermediate transfer belt 10.

FIG. 5 is a diagram showing an example of the image pickup apparatus 21 different from the example shown in FIG.

The image pickup apparatus 21 shown in FIG. 5 extends in the direction indicated by the arrow 10a (see FIG. 4), and extends in the direction indicated by the arrow 10a and the light source 21a for irradiating the surface of the intermediate transfer belt 10 with light. The point that the image pickup unit 21b, such as a line sensor, which is present and photographs the surface of the intermediate transfer belt 10, is separated from each other in the direction indicated by the arrow 10b orthogonal to the direction indicated by the arrow 10a, is the image pickup shown in FIG. It is different from the device 21. The light source 21a irradiates the surface of the intermediate transfer belt 10 with light at a very shallow angle toward the image pickup unit 21b. Therefore, as shown in FIG. 5, when the belt defect 10c is present on the surface of the intermediate transfer belt 10, the light emitted by the light source 21a is obstructed by the belt defect 10c and a shadow is generated. Then, the imaging unit 21b photographs the surface of the intermediate transfer belt 10 at a very shallow angle toward the light source 21a. Therefore, as shown in FIG. 5, when the belt defect 10c is present on the surface of the intermediate transfer belt 10, the imaging unit 21b captures the shadow of the belt defect 10c caused by the irradiation of light by the light source 21a. become.

The belt image obtained by the photographing process of S101 (see FIG. 3) includes a component of a belt defect which is a high frequency component, a noise component which is a high frequency component, and a component of in-plane unevenness which is a low frequency component. Are at least superimposed.

The component of the belt defect is a frequency component corresponding to the size of the belt defect.

The noise component is a frequency component corresponding to minute irregularities generated on the surface of the surface of the intermediate transfer belt.

The components of in-plane unevenness are, for example, a component generated by the peripheral illumination fall of the lens of the image pickup apparatus 21 in the photographing process of S101 (hereinafter referred to as “lens peripheral illumination amount drop component”) and an intermediate transfer belt in the photographing process of S101. A component generated by the gentle deformation of the intermediate transfer belt due to applied tension (hereinafter referred to as "belt deformation component") and a component generated by non-uniform illumination of the intermediate transfer belt in the photographing process of S101 (hereinafter referred to as "belt deformation component"). Hereinafter, it is referred to as “illumination non-uniform component”).

The diagnostic imaging step of S102 (see FIG. 3) is performed by the belt inspection system 30 shown in FIG.

FIG. 6 is a block diagram of the belt inspection system 30 realized by one computer.

The belt inspection system 30 shown in FIG. 6 is an operation unit 31 which is an operation device such as a keyboard or a mouse into which various operations are input, and a display device such as an LCD (Liquid Crystal System) which displays various information. A display unit 32, and a communication unit 33, which is a communication device that directly communicates with an external device via a network such as a LAN (Local Area Network) or the Internet, or directly by wire or wirelessly without a network. It includes a storage unit 34, which is a non-volatile storage device such as a semiconductor memory or an HDD (Hard Disk Drive) for storing various information, and a control unit 35 that controls the entire belt inspection system 30.

The storage unit 34 stores a belt inspection program 34a for detecting a belt defect. The belt inspection program 34a may be installed in the belt inspection system 30 at the manufacturing stage of the belt inspection system 30, for example, a CD (Compact Disk), a DVD (Digital Versaille Disk), a USB (Universal Serial Bus) memory, or the like. It may be additionally installed in the belt inspection system 30 from an external storage medium, or may be additionally installed in the belt inspection system 30 from the network.

The control unit 35 includes, for example, a CPU (Central Processing Unit), a ROM (Read Only Memory) that stores programs and various data, and a RAM (Random) as a memory used as a work area of the CPU of the control unit 35. It is equipped with an Access Memory). The CPU of the control unit 35 executes the program stored in the storage unit 34 or the ROM of the control unit 35.

The control unit 35 has a defect candidate detection unit 35a that detects pixel candidates corresponding to belt defects (hereinafter referred to as “defect candidates”) from the belt image by executing the belt inspection program 34a, and a defect candidate detection unit 35a. A quality determination unit 35b for determining the quality of the intermediate transfer belt based on the defect candidates detected by the above is realized.

In FIG. 6, the belt inspection system 30 is realized by one computer. However, the belt inspection system 30 may be realized by a plurality of computers.

FIG. 7 is a flowchart of the diagnostic imaging process of S102 (see FIG. 3).

As shown in FIG. 7, in the image diagnosis step of S102, the "pretreatment step" (S121) in which the pretreatment for detecting the defect candidate is applied to the belt image, and the pretreatment in the pretreatment step of S121 are performed. A "defect candidate detection step" (S122) for detecting defect candidates based on a belt image, and a "quality determination step" (quality determination step) for determining the quality of an intermediate transfer belt based on the defect candidates detected in the defect candidate detection step of S122. It is composed of S123) and. The pretreatment step of S121 and the defect candidate detection step of S122 are realized by the defect candidate detection unit 35a. The quality determination step of S123 is realized by the quality determination unit 35b.

FIG. 8 is a flowchart of the pretreatment process of S121 (see FIG. 7).

As shown in FIG. 8, the pretreatment step of S121 is an “in-plane unevenness removing step” (S141) for removing in-plane unevenness occurring in the belt image obtained by the photographing step of S101 (see FIG. 3). And "other pretreatment steps" (S142) that execute pretreatment other than removal of in-plane unevenness.

FIG. 9 is a diagram showing an example of an in-plane unevenness removing step of S141 (see FIG. 8).

As shown in FIG. 9, the defect candidate detection unit 35a removes the high frequency component of the belt image obtained by the photographing step of S101 (see FIG. 3) to obtain an image of only the low frequency component (hereinafter, “in-plane”). "Uneven image") is acquired (S161). For example, the defect candidate detection unit 35a may blur the belt image with a smoothing filter such as a moving average filter, a Gaussian filter, or a blur filter.

The band of the high frequency component removed in S161 may be set to such an extent that the component of the belt defect and the component of noise are included without including the component of in-plane unevenness. For example, among the above-mentioned peripheral illumination falloff component, belt deformation component, and illumination non-uniformity component, the component of in-plane unevenness having the highest frequency is usually a belt deformation component, which is a fraction of the width of the intermediate transfer belt. It is an in-plane unevenness with a frequency of about 1/10. Therefore, the high frequency component to be removed in S161 may be, for example, a frequency component as high as about twice or more the frequency of the belt deformation component.

When a smoothing filter is used to remove high-frequency components in S161, there is a large brightness gap between the part where the belt defect exists and the part where the belt defect exists in the vicinity of the part where the belt defect exists. The effect of lightness may remain. Therefore, the defect candidate detection unit 35a converts, for example, the average value of the brightness values of the peripheral pixel groups within the specified distance of the pixel of interest, and calculates and calculates the gap between the brightness value of the pixel of interest and the average value. If the gap exceeds the specified value, the brightness value corresponding to the pixel of interest is calculated by interpolating with a smooth curve from the brightness value of the pixel group in the vicinity of the pixel of interest without using the brightness value of the pixel of interest. You just have to hit it. When the calculated gap is equal to or less than the specified value, the defect candidate detection unit 35a may apply the average value of the brightness value of the pixel of interest and the brightness value of the pixel group in the vicinity of the pixel of interest. By applying such processing to the entire surface of the belt image, the influence of glitch-like signal fluctuations occurring in the part where the presence of belt defects is assumed is removed or reduced, and the high frequency component is removed. Can be done.

FIG. 10A is a schematic diagram showing an example of a brightness value at a specific line of a belt image. FIG. 10B is a schematic diagram showing an example of the brightness value when the high frequency component is removed with respect to the brightness value shown in FIG. 10A.

In FIG. 10A, the difference in brightness value from the peripheral pixels is extremely large at the locations 51a, 51b, and 51c where the belt defect exists. When the influence of the glitch-like signal fluctuation occurring in the portion where the presence of the belt defect is assumed is removed or reduced with respect to the brightness value shown in FIG. 10 (a) and the high frequency component is removed, FIG. 10 (b) is obtained. ). In addition, in FIG. 10 (b), the lightness value shown in FIG. 10 (a) is shown by a thin line for reference.

In the above, the processing is switched based on the gap between the brightness value of the pixel of interest and the average value of the brightness values of the peripheral pixels within the specified distance of the pixel of interest. However, instead of the gap between the brightness value of the attention pixel and the average value, the gap between the brightness value of the attention pixel and the average value is set as the standard of the brightness value of the peripheral pixel group within the specified distance of the attention pixel. Processing may be switched based on the ratio divided by the deviation.

It should be noted that the averaging method at the time of calculation of the above processing may be a load averaging with a weight added to the vicinity. Further, a geometric mean may be used, or an averaging method may be used in which numerical sorting is performed once, upper and lower data are excluded, and only the remaining intermediate data is averaged. Further, when the ratio of the number of pixels of the belt defect to the total number of pixels of the belt image is extremely small and the influence of the brightness value of the pixel of the belt defect on the average value is negligible, a simple addition average may be used. ..

As shown in FIG. 9, the defect candidate detection unit 35a calculates the average value of the brightness values of the in-plane unevenness images acquired in S161 after the processing of S161 (S162).

Here, in the belt image, it is experimentally observed that the component of the belt defect which is a high frequency component and the component of noise which is a high frequency component increase / decrease in correlation with the increase / decrease of the component of in-plane unevenness which is a low frequency component. It has become clear to. That is, in the belt image, the high frequency component increases as the low frequency component increases, and decreases as the low frequency component decreases.

FIG. 11A is a diagram showing a signal in which a low frequency component and a high frequency component that increases or decreases in correlation with the increase or decrease of the low frequency component are superimposed. FIG. 11B is a diagram showing a signal obtained by subtracting a low frequency component from the signal shown in FIG. 11A. FIG. 11 (c) is a diagram showing a signal obtained by dividing the signal shown in FIG. 11 (a) by a low frequency component.

In FIG. 11A, the low frequency component is shown by a broken line. The signal shown in FIG. 11 (b) remains affected by the increase / decrease of the low frequency component in the signal shown in FIG. 11 (a). On the other hand, the signal shown in FIG. 11 (c) does not have the influence of the increase / decrease of the low frequency component in the signal shown in FIG. 11 (a), and only the high frequency component is appropriately extracted from the signal shown in FIG. 11 (a). Can be done.

Therefore, when extracting only the high frequency component from the belt image, it is possible to divide the brightness value of each pixel of the belt image by the brightness value of the corresponding pixel in the low frequency component, instead of simply subtracting the low frequency component from the belt image. preferable.

Therefore, as shown in FIG. 9, the defect candidate detection unit 35a corresponds to the brightness value of each pixel of the belt image obtained by the photographing step of S101 (see FIG. 3) in the in-plane unevenness image acquired in S161. A normalized image is obtained by dividing by the brightness value of the pixel to be used (S163).

Then, the defect candidate detection unit 35a acquires a belt image corrected for in-plane unevenness by multiplying the brightness value of each pixel of the image acquired in S163 by the average value calculated in S162 (S164).

FIG. 12 is a schematic view showing an example of the brightness value when the in-plane unevenness is corrected with respect to the brightness value shown in FIG. 10 (a).

When the in-plane unevenness is corrected with respect to the brightness value shown in FIG. 10 (a), it becomes as shown in FIG. In FIG. 12, the lightness value shown in FIG. 10B is normalized and multiplied by the average value, and is shown by a thin line for reference.

As shown in FIG. 9, the defect candidate detection unit 35a additionally adjusts the brightness of the belt image acquired in S164 after the processing of S164 (S165). The additional adjustment of the brightness in S165 is, for example, changing the gain of the brightness of the belt image acquired in S164 to the gain input from the outside when the gain of the brightness of the belt image is input from the outside as a parameter. When the brightness offset of the belt image is input from the outside as a parameter, the offset input from the outside is given to the brightness of the belt image acquired in S164, or the parameter for gamma correction is input from the outside. , Gamma processing that gives non-linearity to the brightness of the belt image acquired in S164 is performed according to a parameter input from the outside.

For example, when an inspection worker compares the belt image displayed on the monitor with the actual intermediate transfer belt to identify the position of the belt defect in the actual intermediate transfer belt, the variation in in-plane unevenness. If the brightness of the entire belt image is dark, the visibility and usability are not good for the inspection worker. Therefore, the brightness of the belt image acquired in S164 is additionally adjusted in S165. If the brightness of the belt image acquired in S164 is appropriate for the inspection worker, nothing may be done in S165, or the process itself of S165 may not be provided.

FIG. 13 is a flowchart of an example of “other pretreatment steps” of S142 (see FIG. 8).

As shown in FIG. 13, the “other pretreatment step” of S142 is, for example, tilt correction that detects and corrects when the belt image is tilted, rotated, or meandered. Step (S181), a streak removal step (S182) for removing horizontal streaks and vertical streaks that are not defects from the belt image when there are horizontal streaks and vertical streaks on the intermediate transfer belt, and an imaging step of S101 (see FIG. 3). In, the jig of the intermediate transfer belt vibrates, the image pickup system vibrates, the intermediate transfer belt itself vibrates, or the intermediate transfer belt is set incorrectly and the intermediate transfer belt slips or meanders. Visual blur removal step (S183) that corrects motion blur by back calculation and corrects optical blur that occurs when the intermediate transfer belt is out of focus and out of focus during the shooting process of S101. It is composed of a brightness correction step (S184) for correcting the brightness of the belt image in order to ensure the performance and workability.

Regarding the streak removal step of S182, the streaks existing on the intermediate transfer belt include, for example, molding marks during inflation molding in the intermediate transfer belt generation step, intermediate transfer belts during the coating treatment on the surface of the intermediate transfer belt, and the like. These are scratch marks on the intermediate transfer belt with the jig and processing marks on the intermediate transfer belt. In the streak removal step of S182, the defect candidate detection unit 35a subtracts the average value of the lightness values of all the pixels in this line from the lightness value of each pixel in the line corresponding to the streak, for example, in the belt image. Then, the streaks are removed from the belt image by adding the average value of the overall brightness values of the belt image.

In the present embodiment, all the processes in the "other pretreatment steps" of S142 (see FIG. 8) are executed after the in-plane unevenness removing step of S141 (see FIG. 8). However, the defect candidate detection unit 35a may execute at least one process in the “other pretreatment step” before the in-plane unevenness removing step.

FIG. 14 is a diagram showing an example of a defect candidate detection step of S122 (see FIG. 7).

As shown in FIG. 14, when the defect candidate detection unit 35a finishes the pretreatment step of S121 (see FIG. 7), the defect candidate detection unit 35a is a belt image (hereinafter, “preprocessed belt image”) that has been preprocessed in the pretreatment step. The standardization step of standardizing the brightness value of each coordinate of (S201) is executed (S201).

FIG. 15 is a diagram showing an example of the standardization process of S201 (see FIG. 14).

As shown in FIG. 15, the defect candidate detection unit 35a calculates the average value of the brightness values of all the pixels of the preprocessed belt image (S221).

Further, the defect candidate detection unit 35a calculates the standard deviation of all the pixels of the preprocessed belt image (S222).

Next, the defect candidate detection unit 35a standardizes the brightness value of each pixel of the preprocessed belt image (S223). Specifically, the defect candidate detection unit 35a is standardized by dividing the value obtained by subtracting the average value calculated in S221 from the brightness value of each pixel of the preprocessed belt image by the standard deviation calculated in S222. The value, that is, the Z score is calculated. Expressed as an expression, it is as follows.
Zi = (Xi -Xmean) / σ
Zi: Z score of the brightness value of the i-th pixel of the preprocessed belt image
Xi: Brightness value of the i-th pixel of the preprocessed belt image
Xmean: Mean of brightness values of all pixels of preprocessed belt image σ: Standard deviation of brightness values of all pixels of preprocessed belt image

In the process of S221, the defect candidate detection unit 35a may calculate the average value by excluding the brightness value of the pixel of the belt defect portion. Similarly, the defect candidate detection unit 35a may calculate the standard deviation by excluding the brightness value of the pixel of the belt defect portion in the processing of S222.

In the above, the defect candidate detection unit 35a uses the brightness value of the target pixel of the preprocessed belt image and the average value and standard deviation of the brightness values of all the pixels of the preprocessed belt image for preprocessing. The Z score of the brightness value of the target pixel of the finished belt image is calculated. However, the defect candidate detection unit 35a replaces the average value and standard deviation of the brightness values of all the pixels of the preprocessed belt image with the average value of the brightness values of the pixel groups in the vicinity of the target pixels of the preprocessed belt image. And the standard deviation may be used to calculate the Z score of the brightness value of the target pixel of the preprocessed belt image.

As shown in FIG. 14, after the processing of S201, the defect candidate detection unit 35a determines in S201 a pixel candidate corresponding to the “coating defect” (hereinafter referred to as “coating defect candidate”). Each pixel of the preprocessed belt image is binarized based on the calculated Z score and a specific threshold (S202). That is, the defect candidate detection unit 35a classifies pixels having a Z score larger than the threshold value and pixels having a Z score equal to or lower than the threshold value. Here, the brightness of the pixel corresponding to the "painting defect" is usually darker than the average brightness in the belt image. Therefore, the threshold value used in S202 is a negative value such as −6.0, −3.0, and the like. In S202, the defect candidate detection unit 35a may generate a plurality of binarized images by changing the threshold value.

After the processing of S202, the defect candidate detection unit 35a tentatively determines the coating defect candidate based on the binarized image generated in S202 (S203).

Hereinafter, the method of tentatively determining the coating defect candidate in S203 will be specifically described.

FIG. 16 is a diagram showing the probability density of the normal distribution used in the defect candidate detection step shown in FIG.

In the following, an example in which the size of the belt image is 1698 pixels × 4792 pixels will be described. That is, an example in which the number of pixels of the belt image is 8136816 will be described.

For example, the probability that the Z score will be −6.0 or less is 6.07588 × ^10-9 as shown in FIG. Therefore, when the number of pixels of the belt image is 8136816, the expected number of pixels having a Z score of −6.0 or less is 8136816 × 6.07588 × ^10-9 , that is, 0. Therefore, in the binarized image generated with the threshold value of −6.0, the pixels having the Z score equal to or lower than the threshold value may be tentatively determined as a coating defect candidate for each pixel block.

Here, since the intermediate transfer belt is generated by inflation molding, a slight pattern is generated on the surface of the intermediate transfer belt. And this pattern becomes the fluctuation of the brightness in the belt image. In addition, photon noise and dark current noise are also added to the belt image. Further, in the photographing process of S101 (see FIG. 3), the jig of the intermediate transfer belt vibrates, the imaging system vibrates, the intermediate transfer belt itself vibrates, and the method of setting the intermediate transfer belt is bad. The motion blur caused by the intermediate transfer belt slipping or meandering can blur the belt image. Further, in the shooting process of S101, the jig of the intermediate transfer belt vibrates, the image pickup system vibrates, or the intermediate transfer belt itself vibrates, and the image pickup device 21 is optically out of focus. There is a possibility that the belt image will be blurred because it was taken during the focusing control. Due to these various causes, in the belt image, the brightness of the portion of the belt defect may be mixed with the brightness of the portion around the belt defect and averaged. Therefore, even if the pixel has a Z score of more than −6.0, it may actually be a pixel corresponding to a belt defect. Therefore, even if the pixel has a Z score of more than −6.0, it is tentatively determined as a coating defect candidate by satisfying a specific condition.

For example, the probability that the Z score will be -3.0 or less is 0.00443185 as shown in FIG. Therefore, when the number of pixels of the belt image is 8136816, the expected number of pixels having a Z score of −3.0 or less is 8136816 × 0.00443185, that is, 36061. The expected number of pixels in which two pixels having a Z score of −3.0 or less are continuously generated is 36061 × 0.00443185, that is, 159. Further, the expected number of locations where three consecutive pixels having a Z score of −3.0 or less are generated is 36061 × 0.00443185 ² , that is, one. Therefore, in the binarized image generated with the threshold value of -3.0, the pixels at the positions where a plurality of pixels having a Z score equal to or less than the threshold value are adjacent to each other are tentatively determined as a coating defect candidate for each pixel block. Is also good.

In the binarized image generated with the threshold value of -3.0, at least a part of the pixels at the positions where a plurality of pixels having a Z score equal to or less than the threshold value are adjacent to each other is generated with the threshold value of -6.02. In the digitized image, it may be a pixel whose Z score is equal to or less than the threshold value. Therefore, in the binarized image generated with the threshold value of -6.0, the Z score is less than or equal to the threshold value in the pixels whose Z score is less than or equal to the threshold value and in the binarized image generated by the threshold value of -3.0. After combining the pixels at the positions where a plurality of pixels are adjacent to each other, it may be tentatively determined as a coating defect candidate for each pixel block.

After the processing of S203, the defect candidate detection unit 35a calculates various information (hereinafter referred to as “defect information”) for each of the coating defect candidates tentatively determined in S203 (S204). Here, the defect information calculated in S204 includes, for example, the number of pixels, the coordinates of the center of gravity, the major axis, the minor axis, the ratio of the major axis to the minor axis, the extending direction, and the roundness. ..

After the processing of S204, the defect candidate detection unit 35a has a specific threshold value or range indicating the characteristics of the coating defect candidate with respect to at least one type of defect information calculated in S204, and the defect information actually calculated in S204. By comparing the above, the candidate coating defects are determined (S205). That is, the defect candidate detection unit 35a narrows down an appropriate coating defect candidate from the coating defect candidates tentatively determined in S203. For example, the defect candidate detection unit 35a may exclude the coating defect candidates whose number of pixels calculated in S204 is not within a specific range from the coating defect candidates tentatively determined in S203.

After the processing of S205, the defect candidate detection unit 35a captures an image including the defect candidate determined in S205 (hereinafter referred to as “defect candidate image”) from the belt image in a specific size based on the coordinates of the center of gravity of the defect candidate. Acquire (S206). For example, the defect candidate detection unit 35a is a square having a range of the number of pixels r in each of the two directions with the coordinates of the center of gravity as the center in the horizontal direction, and a range of the number of pixels r in each of the directions with the coordinates of the center of gravity as the center in the vertical direction. That is, the defect candidate image may be cut out into a square having a size of (2r + 1) × (2r + 1). When the defect candidate detection unit 35a acquires the defect candidate image when the defect candidate image cannot be cut out from the belt image with a specific size because the coordinates of the center of gravity of the defect candidate are close to the edge of the belt image. , The defect candidate image may be acquired in a specific size by filling the insufficient range with the average value of the brightness values of the belt image. If the end of the intermediate transfer belt is cut off when the intermediate transfer belt is mounted on the image forming apparatus, the defect candidate detecting unit 35a is an intermediate transfer belt to the image forming apparatus among the intermediate transfer belts. When a defect candidate is included in the portion cut off before mounting the defect candidate, it is not necessary to acquire the defect candidate image of the defect candidate.

The defect candidate image acquired in S206 is, for example, an image as shown in FIGS. 1 and 2.

The defect candidate detection unit 35a ends the defect candidate detection step shown in FIG. 14 after the processing of S206.

FIG. 17 is a flowchart of the quality determination process of S123 (see FIG. 7).

As shown in FIG. 17, the quality determination unit 35b determines the type and degree of belt defects (S261). For example, as a method for determining the type and degree of belt defects, a method for determining using a specific criterion (hereinafter referred to as "reference use defect determination method") and a method for determining using artificial intelligence (hereinafter referred to as "reference use defect determination method"). It is considered to be "AI use defect determination method").

First, a standard usage defect determination method will be described.

Hereinafter, an example of a standard use defect determination method for determining the type and degree of defect candidates for coating defects in the quality determination step shown in FIG. 7 will be described.

FIG. 18 shows the difference between the minimum value of the brightness value in the defect candidate image and the average value of the brightness value in the preprocessed belt image obtained from the defect candidate image, and the standard deviation of the brightness value in the preprocessed belt image. It is a figure which shows an example of the relationship with.

In FIG. 18, the vertical axis shows the difference between the minimum value of the brightness value in the defect candidate image and the average value of the brightness value in the preprocessed belt image obtained from the defect candidate image, and the horizontal axis indicates the front. It shows the standard deviation of the lightness value in the processed belt image. In FIG. 18, one point indicates one defect candidate image. In FIG. 18, the black circle is a point showing a defect candidate image corresponding to the belt defect determined by the inspection worker to be the belt defect of the needle-shaped “coating defect”. In FIG. 18, the white circles indicate defect candidate images corresponding to the belt defects determined by the inspection worker to be the belt defects of the fibrous “coating defects”.

As shown in FIG. 18, many of the points indicating the defect candidate image corresponding to the belt defect determined by the inspection worker to be a needle-shaped "coating defect" are the lowest value of the brightness value in the defect candidate image. , A line indicating that the value obtained by dividing the difference from the average value of the lightness values in the preprocessed belt image obtained from this defect candidate image by the standard deviation of the lightness values in the preprocessed belt image is 8.0. It exists on the 61 or in the vicinity of the line 61.

Many of the points showing the defect candidate image corresponding to the belt defect determined by the inspection worker as a fibrous "coating defect" are obtained with the lowest brightness value in the defect candidate image and this defect candidate image. On line 62 or line 62 indicating that the value obtained by dividing the difference from the average value of the lightness value in the preprocessed belt image by the standard deviation of the lightness value in the preprocessed belt image is 6.0. Exists in the vicinity of.

Many of the points showing defect candidate images corresponding to belt defects determined by the inspector to be acceptable "coating defects" are not shown in FIG. 18, but of the brightness values in the defect candidate image. The value obtained by dividing the difference between the minimum value and the average value of the lightness values in the preprocessed belt image obtained from this defect candidate image by the standard deviation of the lightness values in this preprocessed belt image is 3.0. It exists on the line 63 indicating the above, or in the vicinity of the line 63.

From the above, the difference between the minimum value of the brightness value in the defect candidate image and the average value of the brightness value in the preprocessed belt image obtained from this defect candidate image is set as the standard of the brightness value in this preprocessed belt image. The range of values divided by the deviation is appropriately the range for needle-shaped "coating defects", the range for fibrous "coating defects", and the range for acceptable "coating defects". By being set, the difference between the minimum value of the lightness value in the defect candidate image and the average value of the lightness value in the preprocessed belt image obtained from the defect candidate image is set to be the difference between the lightness value in the preprocessed belt image. Based on the value divided by the standard deviation, it is possible to distinguish between needle-like "coating defects", fibrous "coating defects" and acceptable "coating defects".

FIG. 19 is a flowchart of a reference use defect determination method for determining the type and degree of defect candidates for coating defects in the quality determination step of S123 (see FIG. 7).

As shown in FIG. 19, the quality determination unit 35b determines whether or not the defect candidate of the coating defect is detected in the defect candidate detection step of S122 (see FIG. 7) (S281).

When the quality determination unit 35b determines in S281 that the defect candidate of the coating defect is not detected in the defect candidate detection step of S122, the operation shown in FIG. 19 is terminated.

When the quality determination unit 35b determines in S281 that the defect candidate of the coating defect is detected in the defect candidate detection step of S122, the defect candidate for each of the defect candidates of the coating defect detected in the defect candidate detection step of S122 is determined. The difference between the minimum value of the brightness value in the defect candidate image acquired in the detection step and the average value of the brightness value in the preprocessed belt image obtained from this defect candidate image is the difference between the brightness value in this preprocessed belt image. The value divided by the standard deviation is calculated (S282).

Next, the quality determination unit 35b targets one defect candidate that has not yet been targeted in the operation shown in FIG. 19 this time, among the defect candidates of the coating defect detected in the defect candidate detection step of S122 (S283). ).

After the processing of S283, the quality determination unit 35b determines whether or not the value calculated in S282 for the current target defect candidate is included in the range for the needle-shaped “coating defect” (S284). ..

When the quality determination unit 35b determines in S284 that the value calculated in S282 for the defect candidate of the current target is included in the range for the needle-shaped "coating defect", the defect candidate of the current target is needle-shaped. (S285).

When the quality determination unit 35b determines in S284 that the value calculated in S282 for the defect candidate of the current target is not included in the range for the needle-shaped "coating defect", the defect candidate of the current target is determined. It is determined whether or not the value calculated in S282 is included in the range for the fibrous "coating defect" (S286).

When the quality determination unit 35b determines in S286 that the value calculated in S282 for the defect candidate of the current target is included in the range for the fibrous "coating defect", the defect candidate of the current target is fibrous. (S287).

When the quality determination unit 35b determines in S286 that the value calculated in S282 for the defect candidate of the current target is not included in the range for the fibrous "coating defect", the defect candidate of the current target is determined. It is determined whether or not the value calculated in S282 is included in the allowable range for "coating defects" (S288).

When the quality determination unit 35b determines in S288 that the value calculated in S282 for the defect candidate of the current target is included in the allowable range for "coating defect", the defect candidate of the current target can be accepted. It is determined to be a "painting defect" (S289).

When the quality determination unit 35b determines in S288 that the value calculated in S282 for the defect candidate of the current target is not included in the allowable range for "coating defect", the quality determination unit 35b determines the defect candidate of the current target. It is determined that it is an impossible "painting defect" (S290).

When the processing of S285, S287, S289 or S290 is completed, the quality determination unit 35b still targets the defect candidates of the coating defect detected in the defect candidate detection step of S122 in the operation shown in FIG. 19 this time. It is determined whether or not there is no defect candidate (S291).

When the quality determination unit 35b determines in S291 that among the defect candidates of the coating defect detected in the defect candidate detection step of S122, there is a defect candidate that has not yet been targeted in the operation shown in FIG. 19 this time, S283. Execute the process.

When the quality determination unit 35b determines in S291 that among the defect candidates of the coating defect detected in the defect candidate detection step of S122, there is no defect candidate that has not yet been targeted in the operation shown in FIG. 19 this time, FIG. The operation shown in 19 is terminated.

In the standard use defect determination method shown in FIG. 19, the needle-shaped "coating defect" and the fibrous "coating defect" are all treated as "unacceptable" as the degree of defect candidates. However, the quality determination unit 35b determines whether the degree of defect candidates is "acceptable" or "unacceptable" for the needle-shaped "coating defect" and the fibrous "coating defect". It may be possible to determine for each defect candidate. Further, the quality determination unit 35b does not have two stages of "acceptable" and "unacceptable" as the degree of defect candidates for the needle-shaped "coating defect" and the fibrous "coating defect". It may be possible to judge in three or more stages.

In the reference use defect determination method shown in FIG. 19, the minimum value of the brightness value in the defect candidate image acquired in the defect candidate detection step of S122 and the average of the brightness values in the preprocessed belt image obtained from this defect candidate image. The type and degree of defect candidates for the coating defect is determined based on the value obtained by dividing the difference from the value by the standard deviation of the brightness value in this preprocessed belt image. However, the type and degree of defect candidates of the coating defect may be determined based on other criteria.

Next, the AI use defect determination method will be described.

The quality determination unit 35b generates a learning model to be used in the AI use defect determination method before executing the diagnostic imaging step of S102 (see FIG. 3). That is, the inspection worker is made to determine the type and degree of the defect candidate corresponding to the defect candidate image, and the determination result is tagged with the defect candidate image and used as teacher data. Then, the quality determination unit 35b generates a learning model by learning using a plurality of teacher data. The defect candidate image used to generate the learning model may be acquired by the same process as the preprocessing step of S121 and the defect candidate detection step of S122.

Then, in the determination in S261, the quality determination unit 35b uses the learning model generated as described above and the defect candidate image acquired in the defect candidate detection step of S122 to determine the type and degree of defect candidates. judge.

The type and degree of the belt defect determined in S261 may be retained even after the operation shown in FIG. 17 is completed. For example, the type and degree of the belt defect determined in S261 can be used for re-learning the learning model in the AI usage defect determination method if it remains after the end of the operation shown in FIG. Further, when the type and degree of the belt defect determined in S261 are left after the end of the operation shown in FIG. 17, and when the belt defect is found in the intermediate transfer belt after the end of the operation shown in FIG. 17, the belt defect is found. It can be used to determine whether or not the belt defect has been detected in the operation shown in FIG.

As shown in FIG. 17, after the processing of S261, the quality determination unit 35b determines the quality of the intermediate transfer belt based on the type and degree of the belt defect determined in S261 (S262). For example, as a method of judging the quality of the intermediate transfer belt, a method of judging using a specific standard (hereinafter referred to as "standard use quality judgment method") and a method of judging using artificial intelligence (hereinafter referred to as "" AI use quality determination method ").

First, a standard usage quality determination method will be described.

FIG. 20 is a flowchart of a standard use quality determination method in the quality determination process of S123 (see FIG. 7).

As shown in FIG. 20, the quality determination unit 35b determines whether or not the type and degree of belt defects determined in S261 (see FIG. 17) satisfy the criteria for failure (S301). For example, the fail criteria may include the criteria that certain types of belt defects are present in the intermediate transfer belt, or more than a certain degree of certain types of belt defects are specific to the intermediate transfer belt. It may include the criterion that there are more than one.

When the quality determination unit 35b determines that the type and degree of the belt defect determined in S261 satisfy the failure criteria, it determines that the quality of the intermediate transfer belt is rejected (S302), and FIG. 20 The operation shown in is terminated.

When the quality determination unit 35b determines that the type and degree of belt defects determined in S261 do not meet the acceptance criteria, whether or not the type and degree of belt defects determined in S261 meet the acceptance criteria. (S303). For example, the acceptance criteria may include the criteria that none of the belt defects determined in S261 existed.

When the quality determination unit 35b determines in S303 that the type and degree of the belt defect determined in S261 satisfy the acceptance criteria, it determines in S303 that the quality of the intermediate transfer belt is acceptable (S304), and FIG. 20 The operation shown in is terminated.

When the quality determination unit 35b determines in S303 that the type and degree of the belt defect determined in S261 do not meet the acceptance criteria, it determines that the quality of the intermediate transfer belt is unknown (S305), and FIG. 20 The operation shown in is terminated.

Next, the AI usage quality determination method will be described.

Before executing the diagnostic imaging step of S102 (see FIG. 3), a learning model used in the AI usage quality determination method is generated. That is, the inspection worker is made to judge the quality of the intermediate transfer belt whose type, degree and number of the belt defects contained in the intermediate transfer belt itself are known, and the judgment result is the result of the determination of the belt defects contained in the intermediate transfer belt. Tag the combination of type, degree and number to make teacher data. Then, the quality determination unit 35b generates a learning model by learning using a plurality of teacher data. The combination of the type, degree and number of belt defects used to generate the learning model may be acquired by the same process as the pretreatment step of S121, the defect candidate detection step of S122, and the treatment of S261. ..

Then, in the determination in S262, the quality determination unit 35b uses the learning model generated as described above and the combination of the type, degree, and number of belt defects acquired in the process of S261 to perform intermediate transfer. Determine the quality of the belt.

The quality of the intermediate transfer belt determined in S262 may be used for re-learning the learning model in the AI use pass / fail determination method.

As shown in FIG. 17, after the processing of S262, the quality determination unit 35b outputs the quality of the intermediate transfer belt determined in S262 via, for example, the display unit 32 or the communication unit 33 (S263). Here, when the defect candidate image is detected in the defect candidate detection step of S122, the quality determination unit 35b also outputs the defect candidate image detected in the defect candidate detection step of S122 in S263. Further, when the quality determination unit 35b determines the type and degree of the belt defect in S261, the quality determination unit 35b also outputs the type and degree of the belt defect determined in S261 in S263.

The quality determination unit 35b ends the operation shown in FIG. 17 after the processing of S263.

If it is determined that the quality of the intermediate transfer belt is unknown, the quality of the intermediate transfer belt may be requested to an inspection worker.

As described above, the belt inspection system 30 sets the brightness value of each pixel of the belt image and the high frequency component from the belt image in the pretreatment step of applying the pretreatment for detecting the candidate of the belt defect to the belt image. Since the in-plane unevenness is removed from the belt image by dividing by the brightness value of the corresponding pixel in the removed image (S163), it corresponds to the image in which the high frequency component is removed from the belt image from the brightness value of each pixel of the belt image. By subtracting the brightness value of the pixel to be used, the in-plane unevenness can be appropriately removed from the belt image as compared with the case where the in-plane unevenness is removed from the belt image, and as a result, the accuracy of detecting the belt defect can be improved. Can be improved.

Since the belt inspection system 30 detects candidates for belt defects based on the pixels of the belt image whose brightness is equal to or lower than the average of the brightness of the belt image (S202 to S203), the brightness of the belt image is high. It is possible to detect belt defects with a brightness below the average of.

Since the belt inspection system 30 detects candidates for belt defects based on the pixels whose standardized brightness value is equal to or less than a specific negative value among the pixels of the belt image (S202 to S203), the belt defect is detected. It is possible to reduce the influence of individual differences of the intermediate transfer belt on the accuracy of. The belt inspection system 30 may execute the processes of S202 to S203 based on the non-standardized brightness values.

Since the belt inspection system 30 can automate the determination of the quality of the intermediate transfer belt with respect to the intermediate transfer belt whose quality can be determined to pass or fail, the burden on the inspection worker can be reduced.

10 Intermediate transfer belt 10c Belt defect 30 Belt inspection system (computer)
34a Belt inspection program 35a Defect candidate detector

Claims (4)

A defect candidate detecting unit for detecting a candidate for a belt defect as an abnormal portion of the intermediate transfer belt from a belt image as an image of the intermediate transfer belt of the image forming apparatus is provided.
The defect candidate detection unit is
A pretreatment step of applying a pretreatment for detecting the candidate to the belt image, and a pretreatment step.
The defect candidate detection step of detecting the candidate from the belt image subjected to the pretreatment is executed.
In the preprocessing step, the defect candidate detection unit divides the brightness value of each pixel of the belt image by the brightness value of the corresponding pixel in the image from which the high frequency component is removed from the belt image, so that the belt image is used. A belt inspection system characterized by removing in-plane unevenness. The defect candidate detection unit is characterized in that, in the defect candidate detection step, the defect candidate is detected based on the pixels of the belt image having a brightness lower than the average of the brightness of the belt image and having a brightness equal to or less than a specific brightness. The belt inspection system according to claim 1. In the defect candidate detection step, the defect candidate detecting unit subtracts the average value of the brightness values of the belt image from the brightness value of each pixel of the belt image by the standard deviation of the brightness value of the belt image. Calculate the standardized value by dividing,
A claim, wherein a pixel having a brightness equal to or lower than the specific brightness among the pixels of the belt image is a pixel having a standardized value equal to or less than a specific negative value among the pixels of the belt image. The belt inspection system according to 2. A defect candidate detection unit for detecting a candidate for a belt defect as an abnormal portion of the intermediate transfer belt from a belt image as an image of the intermediate transfer belt of the image forming apparatus is realized in a computer.
The defect candidate detection unit is
A pretreatment step of applying a pretreatment for detecting the candidate to the belt image, and a pretreatment step.
The defect candidate detection step of detecting the candidate from the belt image subjected to the pretreatment is executed.
In the preprocessing step, the defect candidate detection unit divides the brightness value of each pixel of the belt image by the brightness value of the corresponding pixel in the image from which the high frequency component is removed from the belt image, so that the belt image is used. A belt inspection program characterized by removing in-plane unevenness.

Images