WO2021010157A1

WO2021010157A1 - Method for detecting orientation non-uniformity defect in retardation film and apparatus for detecting orientation non-uniformity defect in retardation film

Info

Publication number: WO2021010157A1
Application number: PCT/JP2020/025669
Authority: WO
Inventors: 翔太橋本; 増田　修; 崇南條
Original assignee: コニカミノルタ株式会社
Priority date: 2019-07-16
Filing date: 2020-06-30
Publication date: 2021-01-21
Also published as: JPWO2021010157A1; KR20220018599A; TW202122787A; TWI757776B; CN114096833A

Abstract

The present invention addresses the problem of providing a method for detecting an orientation non-uniformity defect in a retardation film and an apparatus for detecting an orientation non-uniformity defect in a retardation film that allow for more reproducible evaluation by performing detection and evaluation of an orientation non-uniformity defect in a retardation film through quantitative evaluation using an optical system. The method for detecting an orientation non-uniformity defect in a retardation film according to the present invention comprises the following steps (1) to (5). (1) Placing the retardation film between two polarizers arranged like crossed Nicols such that the slow axis of the retardation film is rotated and inclined at an angle of -10 to 10° with respect to the absorption axes of the polarizers. (2) Irradiating the retardation film with inspection light through one of the polarizers. (3) Obtaining a luminance image by capturing an image of the retardation film through the other polarizer. (4) Performing a differential operation (difference operation) on the captured luminance image in a direction oblique to the slow axis of the retardation film to enhance edges therein. (5) Binarizing the edge-enhanced luminance image using a predetermined threshold value to obtain bright pixels or dark pixels and detecting orientation non-uniformity from the pixels obtained.

Description

Orientation unevenness defect detection method and orientation unevenness defect detection device for retardation film

The present invention relates to a method for detecting uneven alignment defects in a retardation film and a device for detecting uneven orientation defects. More specifically, by detecting and evaluating the orientation unevenness defect of the retardation film by quantitative evaluation using an optical system, the burden on the operator can be reduced and the reproducibility of the evaluation result can be improved. The present invention relates to an orientation unevenness defect detection method and an orientation unevenness defect detection device used therein.

In recent years, liquid crystal display devices have been widely used for televisions, personal computer monitors, etc. due to their features such as power saving, light weight, and thinness. In addition to the above advantages, the liquid crystal display device has a problem of viewing angle dependence due to the refractive index anisotropy of the liquid crystal cell and the polarizing plate. As a typical method for improving the problem of viewing angle dependence, there is a method using a retardation film.

In the conventional manufacturing of retardation film, it is difficult to completely prevent the occurrence of defects that impair the surface quality of the film even in the manufacturing process that is strictly controlled.

Further, in recent years, as the liquid crystal display device has become thinner, the retardation film used therein has also been required to be thinner, and along with this, orientation unevenness defects that occur due to the manufacturing process have become remarkable. It was.

In order to prevent the film having such defects from flowing out to the market, it is important to inspect the film for defects, and the quality evaluation of orientation unevenness defects has conventionally been performed by visual inspection. However, in such visual qualitative evaluation, the evaluation result greatly depends on the proficiency level of the worker, and the reproducibility is low, and the result may change significantly when the worker changes. In addition, since the number of personnel who can be evaluated with confidence is limited to proficient persons, there has been a problem that the man-hours are biased to some workers and the burden is increased.

As the defect inspection of the optical film, the defect inspection method and the inspection apparatus disclosed in

Patent Documents

1 and 2 are known.

In Patent Document 1, a transparent or translucent film sheet is targeted, a light source and a first polarizing plate are arranged between the light source and the film on one surface of the conveyed film sheet, and linear on the other surface. A second polarizing plate is arranged between the array camera and the linear array camera and the film, and the deviation angle between the first polarizing plate and the second polarizing plate in the polarization direction is within ± 20 °. A method of detecting defect information caused by a foreign substance from a video signal output from a camera has been proposed.

In Patent Document 2, a patterned retardation film used for a 3D television is targeted, and first and second polarizing plates are arranged on a cross Nicol so as to sandwich the film, and the film is formed via the first polarizing plate. It includes a light source that irradiates inspection light, an imaging device that photographs transmitted light of a film through a second polarizing plate to obtain a brightness image, and a defect detection unit that detects defects from the brightness image.

The first and second polarizing plates are in a state where one of the polarization transmission axes is substantially parallel to the optical axis of one of the plurality of retardation regions of the patterned retardation film, and there is no defect. A defect inspection device mainly for foreign matter or the like has been proposed, which adjusts one of the polarization transmission axes so that each brightness of a brightness image becomes the same level in the vicinity of the extinguished state when a film is photographed.

However, in the method proposed in Patent Document 1, the two polarizing plates are too far from the cross Nicol state, and in this state, the amount of light from the light source passing through the polarizing plates is large, and the polarized light generated due to the defect of the film. There is a problem that it becomes difficult to grasp the change in the state and it is not possible to acquire the information on the uneven orientation to be acquired.

Further, in the apparatus proposed in Patent Document 2, since the captured luminance image has the same level in the vicinity of the quenching state, there is a problem that a minute change in the polarization state for detecting the alignment unevenness cannot be captured. .. Furthermore, since it is prioritized to erase the boundary lines of different retardation regions of the patterned retardation film from the captured luminance image by image processing, there is a possibility that even the defective portion originally desired to be extracted may be erased. there were.

Japanese Unexamined Patent Publication No. 11-30591 Japanese Unexamined Patent Publication No. 2013-50393

The present invention has been made in view of the above problems and situations, and the problem to be solved is to detect and evaluate the orientation unevenness defect of the retardation film by quantitative evaluation using an optical system. It is an object of the present invention to provide a method for detecting an orientation unevenness defect of a retardation film capable of reducing the burden and improving the reproducibility of an evaluation result, and an orientation unevenness defect detecting device used therefor.

In order to solve the above problems, the present inventor identifies the cause of the problem by using a device composed of two polarizing plates, a light irradiation device, an imaging device, and a defect detection unit in the process of examining the cause of the problem. By adopting the alignment unevenness defect detection method having the steps of, the burden on the operator is reduced by detecting and evaluating the alignment unevenness defect of the retardation film by quantitative evaluation using an optical system, and the evaluation result is obtained. We have found that a method for detecting uneven orientation defects in a retardation film that can improve reproducibility can be obtained.

That is, the above problem according to the present invention is solved by the following means.

1. 1. This is a method for detecting uneven orientation defects in a retardation film.
A method for detecting orientation unevenness defects in a retardation film, which comprises the following steps (1) to (5).

Step (1): When the retardation film is arranged between the first polarizing plate and the second polarizing plate arranged on the cross Nicol, the slow axis of the retardation film is the first polarizing plate or the second polarized light. Either the first polarizing plate and the second polarizing plate, or the retardation film so as to be within the range of -10 to 10 ° (excluding 0 °) with respect to the absorption axis of either of the plates. The step of rotating and arranging.

Step (2): A step of irradiating the retardation film with inspection light via the first polarizing plate.

Step (3): A step of photographing the retardation film with an imaging device via the second polarizing plate to obtain a luminance image.

Step (4): The captured luminance image is 35 with respect to the slow axis of the retardation film.

A step of obtaining a luminance image in which the edge portion is emphasized by performing differential (difference) processing in a direction within the range of ~ 55 ° or −55 to −35 °.

Step (5): The edge-enhanced luminance image is binarized at a predetermined threshold value to obtain bright pixels above the threshold value and dark pixels below the threshold value, and the pixels are oriented from the bright pixels or dark pixels. Step to detect unevenness.

2. Further, expansion processing is performed on the binarized luminance image in a diagonal direction within a range of 70 to 120 ° with respect to the direction of the differential (difference) processing of the retardation film, and the diagonal direction is changed. The method for detecting uneven orientation defects of a retardation film according to the first item, which comprises the step (6) of performing image processing for emphasizing the components.

3. The method for detecting orientation unevenness defects in a retardation film according to item 2, further comprising a step (7) of performing a shrinkage treatment in the same direction with respect to the luminance image after the expansion treatment.

4. Further, it is provided with the step (8) of extracting the shrink-processed luminance image that satisfies the filter condition defined by the representative length of each pixel region as the orientation unevenness defect component in the direction. The method for detecting an orientation unevenness defect of a retardation film according to the third item, which is characteristic.

5. The angle of view θ of the lens calculated from the distance between the lens tip position of the photographing apparatus and the retardation film and the size of the inspection region on the retardation film in the longitudinal direction is the longitudinal direction of the retardation film. The method for detecting an orientation unevenness defect of a retardation film according to any one of the items 1 to 4, wherein the distance is within 23 °.

6. An alignment unevenness detection device for a retardation film used in the method for detecting an orientation unevenness defect of a retardation film according to any one of the items 1 to 5.
The first polarizing plate and the second polarizing plate arranged on the cross Nicol so as to sandwich the retardation film,
A light source that irradiates the retardation film with inspection light via the first polarizing plate,
An imaging device that photographs the retardation film via the second polarizing plate to obtain a luminance image.
A defect detection unit that detects defects from the luminance image is provided.
The defect detection unit
When the retardation film is arranged between the first polarizing plate and the second polarizing plate arranged on the cross Nicol, the slow axis of the retardation film is either the first polarizing plate or the second polarizing plate. Rotate either the first polarizing plate and the second polarizing plate or the retardation film so as to be within the range of -10 to 10 ° (excluding 0 °) with respect to the absorption axis. An edge that emphasizes the edge portion by differentiating (difference) processing the arranged and photographed brightness image in a direction within the range of 35 to 55 ° or -55 to -35 ° with respect to the slow axis of the retardation film. Detection circuit and
A binarization circuit that binarizes the edge-detected luminance image with a predetermined threshold value and makes each pixel a bright pixel equal to or higher than the threshold value and a dark pixel lower than the threshold value.
An orientation unevenness detection device for a retardation film, which comprises.

7. The defect detection unit
Expansion processing is performed on the binarized luminance image in an oblique direction within a range of 70 to 120 ° with respect to the direction of differentiation (difference) processing of the retardation film, and the component in the oblique direction is performed. The misorientation defect detection device for a retardation film according to item 6, further comprising an image processing circuit that emphasizes.

8. The defect detection unit
The device for detecting orientation unevenness defects in a retardation film according to item 7, further comprising an image processing circuit that performs shrinkage processing in the same direction with respect to the luminance image after performing the expansion treatment.

9. The defect detection unit
Item 6 is characterized in that it has a filtering circuit that extracts the shrink-processed luminance image that satisfies the filter condition defined by the representative length of each pixel region as an orientation unevenness defect component in the direction. The device for detecting uneven orientation of a retardation film according to the above.

10. The angle of view θ of the lens calculated from the distance between the lens tip position of the photographing apparatus and the retardation film and the size of the inspection region on the retardation film in the longitudinal direction is determined in the longitudinal direction of the retardation film. The device for detecting uneven orientation of a retardation film according to any one of items 6 to 9, wherein the temperature is within 23 °.

By the above means of the present invention, by detecting and evaluating the alignment unevenness defect of the retardation film by quantitative evaluation using an optical system, the burden on the operator can be reduced and the reproducibility of the evaluation result can be improved. It is possible to provide a method for detecting an orientation unevenness defect of a retardation film and an orientation unevenness defect detecting apparatus used therefor.

The mechanism of expression or mechanism of action of the effect of the present invention has not been clarified, but it is inferred as follows.

The polarized light is arranged so that the two polarizing plates are arranged on the cross Nicol and the slow axis of the retardation film is within the range of -10 to 10 ° (excluding 0 °) with respect to the absorption axis of the polarizing plate. By rotating and arranging either the plate or the retardation film, it is possible to emphasize the change in the polarization state due to the change in the slow axis and improve the detection accuracy of the defective portion (alignment unevenness defect). Inferred.

Differentiation processing is a process that uses a differential filter to extract edges of an image (extracting a part of an image where the brightness changes suddenly is called edge extraction), but in the case of an image, it is not a continuous value. By taking the difference between the pixel values (“difference” in the present invention), it is treated as an approximation of differentiation.

In order to emphasize the edge of the defective portion, the direction of the differential processing is diagonally specified in the range of 35 to 55 ° or -55 to -35 ° with respect to the slow axis of the retardation film. Normally, when producing a retardation film, the slow axis is oriented in one direction by stretching or the like, but minute uneven orientation due to non-uniformity of the width hand tension applied during stretching or drying is other than the above one direction. As a result, it is presumed that the detection accuracy of the uneven orientation can be improved by facilitating the extraction of the diagonal component of the uneven orientation by the differential processing.

Further, the binarization of the obtained luminance image is for roughly distinguishing the defective portion from the other normal region, and the expansion process and the contraction process further perform the defect in the binarized image. It is presumed that the defect detection accuracy can be further improved by emphasizing the part.

Only the alignment unevenness defect is detected by extracting the shrinkage-processed luminance image that satisfies the filter condition defined by the representative length of each pixel region as the orientation unevenness defect component in the direction. It is presumed that the above image processing can provide an accurate and quantitative method for evaluating orientation unevenness defects.

Schematic diagram showing the outline of the conventional visual evaluation method for uneven orientation defects Schematic diagram illustrating the angle of view of the lens attached to the camera used for inspection Specific example of an image binarized with a predetermined threshold value (original brightness image) Specific example of an image binarized with a predetermined threshold value (binary image) Flow chart of image processing and specific examples of processed images Schematic diagram showing an example of an orientation unevenness defect detecting apparatus for a retardation film of the present invention. Schematic diagram illustrating a defect detection unit having an edge detection circuit, a binarization circuit, an image processing circuit, and a filtering circuit.

The method for detecting orientation unevenness defects in a retardation film of the present invention is characterized by having the above steps (1) to (5). This feature is a technical feature common to or corresponding to the following embodiments.

In an embodiment of the present invention, from the viewpoint of exhibiting the effect of the present invention, the retardation film is further binarized in an oblique direction within a range of 70 to 120 ° with respect to the direction of the differential (difference) processing of the retardation film. Having the step (6) of performing expansion processing on the obtained luminance image and performing image processing for emphasizing the component in the oblique direction is to have an accurate and quantitative alignment unevenness defect by emphasizing the defect portion. It is preferable from the viewpoint of enabling evaluation.

Further, having the step (7) of performing the shrinkage treatment in the same direction with respect to the luminance image after the expansion treatment emphasizes the defective portion and reduces noise, so that the orientation is accurate and quantitative. It is preferable from the viewpoint of enabling uneven defect evaluation.

Further, it is possible to have the step (8) of extracting the shrink-processed luminance image that satisfies the filter condition defined by the representative length of each pixel region as the orientation unevenness defect component in the direction. , It is preferable from the viewpoint of identifying the orientation unevenness and enabling accurate and quantitative alignment unevenness defect evaluation.

The angle of view θ of the lens calculated from the distance between the lens tip position of the photographing apparatus and the retardation film and the size of the inspection region on the retardation film in the longitudinal direction is the longitudinal direction of the retardation film. On the other hand, it is preferably within 23 °.
This is because keeping the angle of view θ within the above range, that is, keeping the distance between the retardation film and the camera at the above angle of view θ reduces the dependence of the inspection light on the incident angle and is the center of the inspection area. This is to make the difference in shading of the image between the part and the corner uniform.

The retardation unevenness detection device of the present invention comprises a first polarizing plate and a second polarizing plate arranged on a cross Nicol so as to sandwich the retardation film, and the retardation film via the first polarizing plate. The defect is provided with a light source that irradiates the inspection light, an imaging device that photographs the retardation film via the second polarizing plate to obtain a luminance image, and a defect detection unit that detects a defect from the luminance image. The detection unit sets the slow axis of the retardation film to -10 to 10 ° (excluding 0 °) with respect to the absorption axis of either the first polarizing plate or the second polarizing plate arranged on the cross Nicol. ), The luminance image taken by rotating and arranging either the first polarizing plate and the second polarizing plate or the retardation film so as to be within the range of) is the slow phase of the retardation film. An edge detection circuit that emphasizes the edge portion by differentiating (difference) processing in a direction within the range of 35 to 55 ° or -55 to -35 ° with respect to the axis, and the luminance image in which the edge is detected are subjected to a predetermined threshold value. It is characterized by having a binarizing circuit for binarizing each pixel into a bright pixel above the threshold and a dark pixel lower than the threshold.

The defect detection unit performs expansion processing on the binarized luminance image in an oblique direction within a range of 70 to 120 ° with respect to the direction of differentiation (difference) processing of the retardation film. It is preferable to have an image processing circuit that emphasizes the components in the oblique direction from the viewpoint of enabling accurate and quantitative evaluation of orientation unevenness defects.

Further, the fact that the defect detection unit has an image processing circuit that performs shrinkage processing in the same direction with respect to the luminance image after performing the expansion processing enables accurate and quantitative orientation unevenness defect evaluation. From the point of view, it is preferable.

Further, a filtering circuit is provided in which the defect detection unit extracts the shrink-processed luminance image that satisfies the filter condition defined by the representative length of each pixel region as an orientation uneven defect component in the direction. It is preferable to have it from the viewpoint of identifying the uneven orientation and enabling accurate and quantitative evaluation of the uneven orientation defect.

The angle of view θ of the lens calculated from the distance between the lens tip position of the photographing apparatus and the retardation film and the size of the inspection region on the retardation film in the longitudinal direction is the longitudinal direction of the retardation film. On the other hand, it is preferable that the temperature is within 23 ° from the viewpoint of efficient work efficiency and quantitative evaluation of orientation unevenness defects.

Hereinafter, the present invention, its constituent elements, and modes and modes for carrying out the present invention will be described in detail. In the present application, "-" is used to mean that the numerical values described before and after the value are included as the lower limit value and the upper limit value.

<< Outline of the method for detecting uneven orientation defects in the retardation film of the present invention >>
The method for detecting orientation unevenness defects in a retardation film of the present invention is characterized by having the following steps (1) to (5).

Step (4): The captured luminance image is differentiated (difference) in a direction within the range of 35 to 55 ° or -55 to -35 ° with respect to the slow axis of the retardation film to emphasize the edge portion. Steps to obtain a brightness image.

Here, the "phase difference film" refers to an optical film that improves the problem of viewing angle dependence of a liquid crystal display device, and is, for example, an optical film having birefringence used in a VA (Visual Element) type liquid crystal display device. is there. The optical film is a transparent resin film having a specific function based on optical characteristics. The optical film is formed by stretching or the like, and if it is normally formed, it has uniform optical characteristics (birefringence, retardation, etc.) in the plane, but in reality, some non-uniformity occurs.

The value of the phase difference of the optical film can generally be defined by the following formula.

That is, the in-plane phase difference Ro and the thickness direction phase difference Rt of the optical film are represented by the following equations (1) and (2).

Ro = (n _{x −} n _y ) × d ・・・ (1)
Rt = {(n _x + n _y ) /2- _nz } × d ・・・ (2)
In the formula, n _x is the refractive index in the slow axis direction in the film surface, n _y is the refractive index in the phase advance axis direction in the film surface, n _z is the refractive index in the thickness direction of the film, and d is the film. Represents the thickness (nm) of.

The "slow phase axis" refers to the axis in which the phase is delayed and the traveling speed of the light is the slowest when the light propagates in the film causing birefringence. That is, it refers to the direction in which the in-plane refractive index is maximized. The "phase-advancing axis" refers to the axis at which the traveling speed of light becomes the fastest, and refers to the direction in which the refractive index in the plane becomes the minimum.

The in-plane slow-phase axis and phase-advance axis of the retardation film can be confirmed by an automatic birefringence meter Axoscan (AxoScan Mueller Matrix Polarimeter: manufactured by Axometrics).

The Ro and Rt of the retardation film can be measured by the following method.

1) Humidity control of the retardation film at 23 ° C. and 55% RH for 24 hours. The average refractive index of this retardation film is measured with an Abbe refractometer, and the thickness d is measured with a commercially available micrometer.

2) The retardation Ro and Rt of the retardation film after humidity control at the measurement wavelength of 550 nm are measured by the automatic birefringence meter Axo Scan (Axo Scan Mueller), respectively.
Measured in an environment of 23 ° C. and 55% RH using a Matrix Polarimator (manufactured by Axometrics).

When the retardation film according to the present invention is used for VA mode, for example, the in-plane retardation Ro measured in an environment with a measurement wavelength of 550 nm, 23 ° C., and 55% RH is within the range of 20 to 120 nm. It is preferably in the range of 30 to 100 nm, and more preferably in the range of 30 to 100 nm. The phase difference Rt in the thickness direction of the optical film is preferably in the range of 70 to 350 nm, and more preferably in the range of 100 to 320 nm.

Since the phase difference is a value that fluctuates depending on the wavelength of the light irradiating the optical film, the environmental conditions (temperature and humidity) for measurement, etc., it depends on the purpose of optical characteristic evaluation and the type of optical film to be evaluated. Therefore, it is desirable to set these conditions so that the phase difference becomes an arbitrary value.

Here, the "orientation angle" refers to an angle formed by the orientation direction in which the molecules forming the retardation film are oriented with respect to a predetermined reference direction, and usually, the in-plane slow axis of the retardation film is a predetermined reference. Matches the angle made with respect to the direction. The predetermined reference direction is usually the width direction of the retardation film and the direction in which the film is stretched. The "alignment unevenness defect" according to the present invention means that the direction of the slow axis deviates from the reference direction at each sampled portion in the film plane (for example, when a plurality of samples are sampled along the width direction). Defects with unevenness.

A "polarizing plate" is a type of polarizing element, and is usually composed of a polarizer and a protective film. The polarizer is used to change the vibration direction of the incident linearly polarized light by rotating the transmission axis direction with respect to the linearly polarized light having a specific vibration direction extracted from natural light (randomly polarized light). A "polarizer" is an element that allows only light on a plane of polarization in a certain direction to pass through, and is a polyvinyl alcohol-based polarizing film. The polyvinyl alcohol-based polarizing film includes a polyvinyl alcohol-based film dyed with iodine and a film obtained by dyeing a dichroic dye.

"Absorption axis" refers to the direction of the axis in which the polarizer absorbs the vibration of light, and usually refers to the angle that coincides with the stretching direction of the polarizer.

The "first polarizing plate and the second polarizing plate arranged on the cross Nicol" means that the "absorption axis" of the polarizer is orthogonal to the first polarizing plate and the second polarizing plate in the 90 ° direction. Say.

Hereinafter, the present invention, its constituent elements, and modes and modes for carrying out the present invention will be described. In the present application, "-" is used to mean that the numerical values described before and after the value are included as the lower limit value and the upper limit value.

[1] Configuration of Method for Detecting Orientation Uneven Defects in Displacement Film The method for detecting orientation unevenness defects in a retardation film of the present invention includes the steps (1) to (5), and more preferably the step (5). It is preferable to have step (6), step (7) and step (8) following.

First, the outline of the conventional visual evaluation method for orientation unevenness defects will be described.

FIG. 1 is a schematic diagram showing an outline of a conventional visual evaluation method 10 for uneven orientation defects.

In the evaluation of the alignment unevenness defect of the retardation film 1, the irradiation light L emitted from the flat LED illumination 2 at the lower part is transmitted in the order of the first polarizing plate 3, the retardation film 1 and the second polarizing plate 4, and the present invention is performed. This is a method of qualitatively observing the alignment unevenness defect of the retardation film 1 visually 5. At this time, since the absorption axes of the first polarizing plate 3 and the second polarizing plate 4 are arranged so as to form a cross Nicol, the irradiation light L is not transmitted as it is and becomes a dark part, but the evaluator makes a phase difference. By slightly rotating the film 1 from the absorption axis direction of the polarizing plate (inclining in the inclination direction 6 of the arrow in the figure), the irradiation light L is polarized according to the direction of the slow axis of the retardation film and transmitted light. If there is an orientation unevenness defect in visual inspection 5, it is observed as a dark region.

Therefore, the evaluation is personal and qualitative because the evaluator's proficiency in how to handle the retardation film and the ranking based on the evaluator's experience for orientation unevenness defects affect the evaluation result. This is an evaluation method.

Since the method for detecting orientation unevenness defects in the retardation film of the present invention is a method for evaluating the defects in a certain step using an optical evaluation device, the evaluation is not personal and quantitative evaluation. Can be provided.

Hereinafter, the details of each step of the method for detecting orientation unevenness defects of the retardation film of the present invention will be described.

When the film is placed between two polarizing plates (first polarizing plate and second polarizing plate) in the cross Nicol state, the slow axis of the film is placed so as to be the same as the absorption axis of the polarizing plate. However, light does not pass through and nothing can be observed. When the film is rotated slightly with respect to the absorption axis of the polarizing plate, the polarization state of the light passing through the film changes depending on the slow axis of the retardation film (the direction of polarization changes slightly), so that the second polarizing plate Light will reach through.

The uneven orientation defect, which is the subject of this inspection, has a small change in brightness, so if the polarizing plate or retardation film is rotated too much, the light that reaches the camera, which is the imaging device, is too strong to observe. As a result of the examination, the first polarizing plate is set so that the slow axis of the retardation film is within the range of -10 to 10 ° with respect to the absorption axis of either the first polarizing plate or the second polarizing plate. And, when either the second polarizing plate or the retardation film is rotated, the retardation unevenness can be observed most clearly, so this is specified in this range.

In an actual operation, in the orientation unevenness defect detecting device of FIG. 5, which will be described later, the first polarizing plate and the second polarizing plate are rotated without rotating the retardation film, and the absorption axis of the polarizing plate is set. It is preferable to incline with respect to the slow axis of the retardation film.

As the lighting device that irradiates the inspection light, it is necessary to transmit the inspection light in the order of the first polarizing plate, the retardation film, and the second polarizing plate, and a sufficient area is required to photograph the inspection light, and the brightness is as high as possible. It is preferably uniform.

The light source of the lighting device is not particularly limited, but is a fiber-type light source using a heavy hydrogen lamp, a halogen lamp, or an LED (Light Emitting Diet) lamp as a light source, or a fluorescent lamp, an LED, or an LED (Organic Light Emitting Diet). ) Can be used as a surface light source. It is preferably a surface light source (flat surface light source), and it is preferable to use flat LED illumination.

The amount of light for inspection is managed on the inspection machine control software LabVIEW (manufactured by (National Instruments)), and the brightness calculation area (for example, a rectangular area of about 80 mm × 60 mm to 400 mm × 300 mm) set on the LabVIEW It is preferable to measure the in-plane average brightness and adjust the amount of light of the light source so that the brightness value falls within the range of 128 ± 10 in order to obtain a sufficient evaluation brightness image.
Further, the brightness calculation region is affected by the incident angle of the inspection region, and a phenomenon occurs in which the brightness calculation region becomes darker toward the end of the inspection region. If an attempt is made to adjust the in-plane average brightness to 128 ± 10 including the dark part at the end, the central part becomes excessively bright and the essential unevenness cannot be detected. Therefore, the brightness calculation area. Is preferably set narrower than the inspection area. For example, when the inspection area is 400 mm in length × 300 mm in width, the brightness calculation area is preferably set to 90 mm in length × 80 mm in width.

The photographing device is not particularly limited, but an area type or line sensor type CCD camera can be used. If the inspection area range of the retardation film can be covered within a few shots, it is preferable to use the area type, and if it is necessary to shoot a wider range, the line sensor type is preferable in terms of shooting time and accuracy. ..

For the camera as the photographing device, for example, the model acA2500-14 gm manufactured by Basler AG, and the lens model H6X8-1.0-II manufactured by APACECOM can be used.

The inspection area is determined by the angle of view of the lens mounted on the camera, the size of the polarizing plate and the size of the retardation film, and the distance between the retardation film and the lens mounted on the camera. If the imaging device is configured so that the sample can be moved in the width direction, it is possible to measure a sample of any width by performing the inspection multiple times.

For example, when the inspection area is 400 mm in length and 300 mm in width for a sample having a length of 450 mm and a width of 2500 mm as a retardation film, the inspection may be performed by a plurality of imaging shots. The size in the longitudinal direction is fixed by the inspection area of the device. The size of the width is the maximum width that is generally used as a retardation film.

Further, the angle of view θ of the lens mounted on the camera is determined in detail from the distance between the lens tip position of the photographing apparatus and the retardation film and the size of the inspection region on the retardation film in the longitudinal direction. Although it is calculated, it is preferable that the angle of view θ is within 23 ° with respect to the longitudinal direction of the retardation film from the viewpoint of enabling work efficiency and quantitative alignment unevenness defect evaluation.

FIG. 2 is a schematic view illustrating an angle of view θ of a lens mounted on a camera used for inspection.
The distance WD (abbreviation of working distance, the distance from the tip of the lens to the position where the subject is in focus) between the tip position of the lens 56 and the retardation film 51 is determined by, for example, the alignment unevenness defect inspection apparatus according to the present invention. Assuming that the WD is 1150 mm, the imaging region of the camera is a lens when the longitudinal length 51a of the retardation film 51 is 450 mm and the width length 51b is 375 mm with respect to the transport direction 51c of the retardation film 51. The angle of view θ of is 22.14 °. When the inspection area has a length of 400 mm and a width of 300 mm, θ is 19.73 °. Both can be calculated.

The purpose of general differential processing is to emphasize the edge portion of the captured luminance image. Differentiation processing is a process that uses a differential filter to extract edges of an image (extracting a part of an image where the brightness changes suddenly is called edge extraction), but in the case of an image, it is not a continuous value. By taking the difference between the pixel values (“difference” in the present invention), it is treated as an approximation of differentiation. Differentiation processing is performed at the examination stage by differentiating a matrix called a kernel (also called a matrix) in a direction within a range of 35 to 55 ° and a kernel in a direction within a range of -55 to -35 °. It is preferable to find and use the kernel of the time.

The edge refers to a portion of the image that is divided into light and dark areas. When emphasizing the edge, it is ideal to process it in the direction orthogonal to the direction in which the boundary between light and dark extends. Orientation unevenness defects appear along the direction of about 45 ° with respect to the longitudinal direction (or the width direction), but there is some variation in this angle, and in order to correspond to this, the orientation of the differential processing is set to the retardation film. It is preferable to have a width in the range of 35 to 55 ° with respect to the slow axis of.

The differential processing software to be used is not limited, but it is preferable to use the analysis software NI vision (manufactured by National Instruments).

Step (5): The edge-enhanced luminance image is binarized with a predetermined binarization software and a threshold value to obtain bright pixels above the threshold value and dark pixels below the threshold value, and the bright pixels are obtained. Alternatively, a step of detecting uneven orientation from dark pixels.

In the captured image, the uneven orientation defect is observed as a dark region. Therefore, in the case of binarization, since it is desired to leave a group of elements that appear dark in the image, a threshold value is set and only elements having a brightness lower than the threshold value are left.

At this time, the threshold value to be set is determined while confirming that the uneven part that can be visually confirmed remains clearly. That is, the threshold value may be determined empirically.

Specifically, it is preferable to perform binarization by the following method.

[A] The captured luminance image is read into a personal computer using the analysis software NI vision (manufactured by National Instruments). Since the luminance image changes by adjusting the focus, contrast, and brightness, it is preferable not to make it artificial.

[B] Set the definition of black and white.

In the analysis software NI vision, when a check mark is put in the Black Backg round, the brightness value 0 is expressed in black and the brightness value 255 is expressed in white. If no check mark is put in the Black Backg round, the brightness value 0 is represented by white and the brightness value 255 is represented by black.

[C] Image noise removal Shading processing is performed. The shading process removes noise by averaging the brightness in the image.

[D] Apply a bandpass filter.

For example, a bandpass filter value of 20 to 100 is recommended. Since this set value depends on the initial luminance image, it is preferable to set it appropriately and optimally.

[E] Binarize the luminance image.

Set the threshold to 8 bits in the settings. The threshold value is preferably adjusted according to the analysis software NI vision (manufactured by National Instruments).

Since this threshold value changes depending on the contrast of the image, it is preferable that the evaluator sets it every time instead of fixing it.

When the threshold is decided, make a black and white image. Binarization is performed with a predetermined threshold value, each pixel is obtained as a dark pixel above the threshold value and a bright pixel lower than the threshold value, and uneven orientation is detected from the bright pixel or the dark pixel.

FIG. 3 is a specific example of an image binarized with a predetermined threshold value.

FIG. 3A is the original luminance image taken, and the dark pixel portion that looks dark in the image is a candidate for the orientation unevenness defect. FIG. 3B shows a threshold value determined and binarized as dark pixels so that candidates for the uneven orientation defect remain clearly.

The method for detecting uneven orientation defects of the present invention preferably further includes the following steps (6), (7), and (8).

Step (6): The binarized luminance image is expanded in an oblique direction within a range of 70 to 120 ° with respect to the direction of the differential (difference) processing of the retardation film. A step of performing image processing that emphasizes components in the diagonal direction.

Since a lot of noise other than the alignment defect unevenness element remains in the image immediately after binarization, expansion processing is performed to emphasize the defective part, and the expansion processing makes it easier to reduce the noise. Here, the "expansion process" is a process in which a binary black-and-white image is generally processed, and if there is even one white pixel around the pixel of interest, the process is replaced with white. That is. On the contrary, the "shrinkage process" described later is called "erosion", which is a process of replacing even one pixel with black pixels in the periphery.

In order to remove noise while leaving the alignment defect unevenness, it is effective to process along the direction in which the unevenness exists. If this is expressed with reference to the direction of the differential processing, it will be processed in the direction orthogonal to the differential processing. This is because it is ideal that the direction of the differential processing is orthogonal to the direction in which the unevenness exists. However, since the direction of unevenness also varies, in order to deal with this variation, the angle should be widened and the expansion process should be performed diagonally within the range of 70 to 120 ° with respect to the direction of the differential (difference) process. It is preferable to do so.

Step (7): A step of performing a contraction process in the same direction with respect to the luminance image after performing the expansion process.

Performing the contraction process after the expansion process is also a process for removing noise as much as possible. Similarly, since the direction of the differential process is performed in the direction orthogonal to the direction in which the unevenness exists, the angle in the direction in which the unevenness exists. It is preferable that the shrinkage treatment is performed in an oblique direction within a range of 70 to 120 ° with respect to the direction of the differential (difference) processing.

The expansion treatment and the contraction treatment can be performed using the analysis software NI vision (manufactured by National Instruments).

Step (8): A step of extracting a shrink-processed luminance image that satisfies the filter condition defined by the representative length of each pixel region as an orientation unevenness defect component in the direction.

This step is also referred to as filtering or filtering, and the "filter condition defined by the representative length of each pixel region" is a condition that is empirically required in detecting an orientation unevenness defect.

From the binarized luminance image, for example, those whose shape is close to a perfect circle are removed (filter processing 1), elements with a short length are removed (filter processing 2), and the area is outside the specified range. (Filtering process 3) is performed, and those satisfying the filter condition defined by the representative length of each pixel area are detected as alignment unevenness defects.

Specifically, for example, when the filter process 1 is the quotient of Haywood circular factor H: the circumference of the pixel region divided by the circumference of the same area, the pixels are within the range of 1.5 ≦ H ≦ 5. It is preferable to leave only the area. The Haywood circular factor H is 1 in the case of a perfect circle.

Further, as the filtering process 2, it is preferable to remove the pixel region having a specific length or less on the long side of the uneven orientation.

As the filtering process 2, when the expansion factor L: the maximum ferret diameter of the pixel region is divided by the short side (ferret) of the value rectangle, only the pixel region within the range of 4 ≦ L ≦ 13 is left. Therefore, it is preferable to detect alignment unevenness defects.

The filter processing software used in the filter processing is not limited, but can be performed by using the analysis software NI vision (manufactured by National Instruments).

FIG. 4 shows a flowchart of image processing and a specific example of processed images.

(P1) Take a retardation film and acquire a luminance image.
(P2) Luminance image shading processing (equalizing the luminance in the image to remove noise),
(P3) Emphasis of edge portion by differentiation (difference) processing of luminance image (P4) Binarization of luminance image,
(P5) Binary image enhancement processing (expansion and contraction processing),
(P6) Filtering to a binarized image 1 (removes elements whose shape is close to a perfect circle),
(P7) Filtering to binary image 2 (removes short elements),
(P8) Filtering 3 (removing elements outside the specified area) to the binarized image,
(P9) Only orientation unevenness defects are detected from the binarized image by filtering.

[2] Alignment unevenness defect detection device for retardation film The alignment unevenness defect detection device for retardation film used in the method for detecting alignment unevenness defect of retardation film of the present invention is arranged on a cross Nicol so as to sandwich the retardation film. The first and second polarizing plates, the light source for irradiating the retardation film with inspection light via the first polarizing plate, and the retardation film taken through the second polarizing plate to obtain brightness. A photographing device for obtaining an image and a defect detecting unit for detecting defects from the luminance image are provided, and the defect detecting unit is an absorption shaft of either the first polarizing plate or the second polarizing plate arranged on the cross Nicol. On the other hand, the first and second polarizing plates or the retardation film so that the slow axis of the retardation film is within the range of -10 to 10 ° (excluding 0 °). The luminance image taken by rotating and arranging any of the above is subjected to differentiation (difference) processing in a direction within the range of 35 to 55 ° or -55 to -35 ° with respect to the slow axis of the retardation film. An edge detection circuit that emphasizes the edge portion, and a binarization circuit that binarizes the edge-detected luminance image with a predetermined threshold and makes each pixel a bright pixel above the threshold and a dark pixel lower than the threshold. It is characterized by having.

FIG. 5 is a schematic view showing an example of an orientation unevenness defect detecting device for a retardation film of the present invention.

The retardation unevenness defect detecting device 50 of the retardation film of the present invention transmits the irradiation light L emitted from the flat LED illumination 52 at the lower part to the light shielding plate 53, the first polarizing plate 54, the retardation film 51, the light shielding plate 53, and the light shielding plate 53. This is a device that transmits light in the order of the second polarizing plate 55 and photographs the alignment unevenness defect of the retardation film 51 with a camera 57 equipped with a lens 56. At this time, since the absorption axes of the first polarizing plate 53 and the second polarizing plate 54 are arranged so as to form a cross Nicol, the irradiation light L is not transmitted as it is and becomes a dark portion, but the retardation film 51 is used. By slightly rotating (tilting) the polarizing plate from the absorption axis direction, the irradiation light L is polarized according to the direction of the slow axis of the retardation film 51, and the uneven orientation is captured as a dark region in the captured image.

In the alignment unevenness defect detecting device 50, the retardation film 51 is not rotated, and the first polarizing plate 54 and the second polarizing plate 55 are absorbed by the polarizing plate with respect to the slow axis of the retardation film. It is preferable to have a mechanism (not shown) that rotates the shaft so as to be in the range of −10 to 10 °.

The dimensions of each part are not particularly limited, but for example, as the evaluation retardation film, a strip-shaped sample cut into 450 mm in the longitudinal direction and 2000 mm in the width direction is used. The photographing device determines the angle of view θ of the lens calculated from the distance between the lens tip position of the photographing device and the retardation film and the size of the inspection region on the retardation film in the longitudinal direction. It is preferable that the temperature is within 23 ° with respect to the longitudinal direction of the above. For example, in order to inspect a region of 400 mm in the longitudinal direction and 300 mm in the width direction, it is a viewpoint of inspection efficiency to assemble an inspection apparatus having the configuration shown in FIG. Is preferable.

The light source of the lighting device is not particularly limited, but as described above, a heavy hydrogen lamp, a halogen lamp, a fiber type light source using an LED lamp, or a surface light source using a fluorescent lamp, an LED, or an LED is used. Can be used.

The photographing device is not particularly limited, but an area type or line sensor type CCD camera can be used. If the measurement target range of the retardation film can be covered within a few shots, it is preferable to use the area type, and if it is necessary to shoot a wider range, the line sensor type is preferable in terms of shooting time and accuracy. ..

The specifications of the camera and lens are as described above.

Further, the defect detection unit performs expansion processing in an oblique direction within a range of 70 to 120 ° with respect to the binarized luminance image with respect to the direction of differentiation (difference) processing of the retardation film. It is preferable to have an image processing circuit that emphasizes the components in the oblique direction.

Further, it is preferable that the defect detection unit has an image processing circuit that performs shrinkage processing in the same direction with respect to the luminance image after performing the expansion treatment.

Further, a filtering circuit is provided in which the defect detection unit extracts the shrink-processed luminance image that satisfies the filter condition defined by the representative length of each pixel region as the orientation unevenness defect component in the direction. It is preferable to have.

FIG. 6 is a schematic diagram illustrating a defect detection unit having an edge detection circuit, a binarization circuit, an image processing circuit, and a filtering circuit.

[Defect detector]
Each configuration of the defect detection unit 100 will be described.

As shown in FIG. 6, the defect detection unit 100 includes a control unit 101, a recording unit 102, a communication unit 103, a data processing unit 104, an operation display unit 105, and the like, and each unit is connected to each other so as to be communicable by a bus 106. There is.

The control unit 101 includes a CPU (Central Processing Unit) 101a that comprehensively controls the operation of the defect detection unit 100, and a RAM (Random) that functions as a work memory for temporarily storing various data when the CPU 101a executes a program. It includes an Access Memory) 101b, a program read and executed by the CPU 101a, a program memory 101c in which fixed data is stored, and the like. The program memory 101c is composed of a ROM or the like.

The recording unit 102 stores and records image data that has been photographed and image-processed, data of various threshold values used in the image processing circuit, and irradiation condition data of the lighting device 52.

The communication unit 103 is provided with a communication interface such as a network I / F, and transmits the measurement conditions input from the operation display unit 105 to the photographing adjustment device 120 via a network such as an intranet. In addition, the communication unit 103 receives the image data sent from the photographing device 130.

The data processing unit 104 performs image processing from the luminance image data received by the communication unit 103 and photographed by the photographing device 130 (lens 56 and camera 57).

The data processing unit 104 has various processing circuits used in steps (4) to (8) according to the present invention.

The data processing unit 104 performs differential (difference) processing on the shading processing circuit 104a in a direction within the range of 35 to 55 ° or -55 to -35 ° with respect to the slow axis of the retardation film to emphasize the edge portion. The edge detection circuit 104b (step (4)) and the luminance image whose edge is detected are binarized at a predetermined threshold value, and each pixel is converted into a bright pixel above the threshold value and a dark pixel lower than the threshold value. The conversion circuit 104c (step (5)), the image processing circuit 104d (step (6) and step (7)) that performs expansion processing or contraction processing on the luminance image, and the luminance image that has been contracted. It also has a filtering circuit 104e (step (8)) that extracts those satisfying the filter condition defined by the representative length of each pixel region as the alignment unevenness defect component in the direction.

The operation display unit 105 may be composed of, for example, an LCD (Liquid Crystal Display), a touch panel provided so as to cover the LCD, various switches and buttons, a numeric keypad, an operation key group, and the like (not shown). The operation display unit 105 receives an instruction from the user and outputs the operation signal to the control unit 101. Further, the operation display unit 105 displays on the LCD an operation screen for displaying various setting screens for inputting various operation instructions and setting information, various processing results, and the like according to the display signal output from the control unit 101.

[External output device]
The data processing device 100 may include an external output device 150 that is communicably connected to the data processing device 100. The external output device 150 may be a general PC (Personal Computer), an image forming device, or the like. Further, the external output device 150 may function as an operation display unit instead of the operation display unit 105 of the data processing device 100.

Hereinafter, the present invention will be specifically described with reference to examples, but the present invention is not limited thereto. In the examples, the indication of "parts" or "%" is used, but unless otherwise specified, it indicates "parts by mass" or "% by mass".

[Example 1]
<Manufacturing of retardation film for evaluation>
An evaluation retardation film 101 was produced according to the following method.

(Preparation of fine particle dispersion diluent)
10 parts by mass of Aerosil R812 (manufactured by Nippon Aerosil Co., Ltd., primary average particle size: 7 nm, apparent specific gravity 50 g / L) and 90 parts by mass of ethanol are stirred and mixed with a dissolver for 30 minutes, and then Manton Goulin, which is a high-pressure disperser. To prepare a fine particle dispersion liquid by dispersing using.

88 parts by mass of dichloromethane was added to the obtained fine particle dispersion while stirring, and the mixture was stirred and mixed with a dissolver for 30 minutes to dilute. The obtained solution was filtered through a polypropylene wind cartridge filter TCW-PPS-1N manufactured by Advantech Toyo Co., Ltd. to obtain a fine particle dispersion diluent.

(Preparation of in-line additive solution)
To 100 parts by mass of dichloromethane, 36 parts by mass of the prepared fine particle dispersion diluted solution was added with stirring, and after further stirring for 30 minutes, 6 parts by mass of cellulose acetate propionate (CAP: acetyl group substitution degree 2.00). , Propionate substitution degree 0.60, weight average molecular weight 270,000) was added with stirring, and the mixture was further stirred for 60 minutes. The obtained solution was filtered through Finemet NF manufactured by Nippon Seisen Co., Ltd. to obtain an in-line additive solution. The filter medium used had a nominal filtration accuracy of 20 μm.

(Preparation of doping)
The following components were placed in a closed container and completely dissolved while heating and stirring. The obtained solution was filtered through a filter equipped with a leaf disc filter at a temperature of 50 ° C. to obtain a main doping. The filter medium used had a nominal filtration accuracy of 20 μm.

<Composition of main doping>
Cellulose acetate propionate (CAP: acetyl group substitution degree 2.00, propionate substitution degree 0.60, weight average molecular weight 270,000)
100 parts by mass Plasticizer: Sugar ester; BzSc (Benzyl sucrose: Average transesterification degree = 5.5)
12 parts by mass Dichloromethane 430 parts by mass Methanol 11 parts by mass 100 parts by mass Main doping 1 and 2.5 parts by mass of in-line additive liquid are sufficiently mixed with an in-line mixer (Toray static in-tube mixer Hi-Mixer, SWJ). Mixing gave a dope.

(Film formation process)
The obtained dope was uniformly cast on a stainless band support using a belt casting device under the conditions of a doping liquid temperature of 35 ° C. and a width of 1950 mm and a final film thickness of 40 μm. On the stainless band support, the organic solvent in the obtained doping film was evaporated until the residual solvent amount reached 100% by mass to form a web, and then the web was peeled off from the stainless band support. The obtained web was pre-dried at 110 ° C. for another 5 minutes to adjust the residual solvent amount to 10% by mass, and then the web was tentered at 160 ° C. to 1.4 times the original width in the TD direction. It was stretched to give the following predetermined phase difference. The stretching rate was 300% / min.

After stretching with a tenter, relaxation was performed at 130 ° C. for 1 minute, and then drying was completed while transporting the drying zone with a large number of rollers. The drying temperature was 130 ° C., and the transport tension was 100 N / m. The obtained film was slit to a width of 2000 mm, both ends of the film were knurled with a width of 10 mm and a height of 5 μm, wound on a core having an inner diameter of 15.24 cm with an initial tension of 220 N / m and a final tension of 110 N / m, and a length of 4000 m. , An evaluation retardation film 101 having a dry film thickness of 40 μm was obtained.
The retardation of the retardation film 101 was Ro: 50 nm and Rt: 120 nm as a result of measurement by the above-mentioned measuring method.

<Evaluation of uneven orientation defects>
Using the alignment unevenness defect evaluation apparatus shown in FIG. 5, the orientation unevenness defects of the above-mentioned produced retardation film were detected and evaluated by the following steps (1) to (8).

The evaluation retardation film was cut into strip-shaped samples of 450 mm in the longitudinal direction and 2000 mm in the width direction. The angle of view θ of the lens calculated from the distance between the lens tip position of the photographing apparatus and the retardation film is 1150 mm and the size of the inspection region on the retardation film in the longitudinal direction is set to 20 °. As a result, imaging was performed a plurality of times in the inspection area of 400 mm in the longitudinal direction and 300 mm in the width direction.

Step (1): When the retardation film is arranged between the first polarizing plate and the second polarizing plate arranged on the cross Nicol, the retardation film is delayed with respect to the absorption axis of the polarizer of the first polarizing plate. A step of rotating and arranging the first polarizing plate and the second polarizing plate so that the phase axis is −7 °.

The angles shown in Table I are negative for counterclockwise and positive for clockwise.

Step (2): A step of irradiating the retardation film with inspection light from flat LED illumination via the first polarizing plate.

Step (3): A step of photographing the retardation film with a photographing device (camera) via the second polarizing plate to obtain a luminance image.

Step (4): A step of obtaining a luminance image in which the edge portion is emphasized by differentiating (difference) the captured luminance image in the direction of 40 ° with respect to the slow axis of the retardation film.

Step (6): Expansion processing is performed on the binarized luminance image in a diagonal direction of 90 ° with respect to the direction of differentiation (difference) processing of the retardation film, and the components in the diagonal direction are extracted. Steps to perform image processing to emphasize.

(Filtering 1)
Haywood circular factor H: When the quotient is obtained by dividing the circumference of the pixel region by the circumference of the same area, only the pixel region within the range of 1.5 ≦ H ≦ 5 is left.

(Filtering 2)
Stretch factor L: Only the pixel region within the range of 4 ≦ L ≦ 13 was left when the maximum ferret diameter of the pixel region was taken as the short side of the value rectangle (quotient divided by the ferret).

[Examples 2 to 8]
In the evaluation conditions of Example 1, the device conditions θ, θf and θp (each representing an angle) and the image processing conditions (step (4), step (5), step (6), step (7) shown in Table I. ), Step (8) -1: Filtering 1 and (8) -2: Filtering -2, and the same evaluation was performed to make Examples 2 to 8.
Here, θ: the angle of view of the lens mounted on the camera, θf: the angle formed by the retard phase axis of the retardation film and the absorption axis of the polarizing plate, and θp: the angle formed by the absorption axes of the two polarizing plates. Represents each.

[Comparative Examples 1 to 3]
Using the orientation unevenness defect evaluation device shown in FIG. 1, three evaluators were evaluated: 1 year or more of evaluation experience (visual 1), 6 months or more and less than 1 year of evaluation experience (visual 2), and less than 6 months of evaluation experience (visual 3). Evaluation of Comparative Examples 1 to 3 was carried out.

[Comparative Example 4]
Evaluation was carried out in accordance with the examples described in paragraphs [0113] to [0122] of JP-A-11-30591.

[Comparative Example 5]
Evaluation was carried out in accordance with Example 1 described in paragraphs [0998] to [0099] of JP2013-50393.

≪Evaluation≫
(1) Complicity The complexity was judged by the time taken to evaluate one sample.

⊚: less than 1 minute 〇: 1 minute or more and less than 5 minutes Δ: 5 minutes or more and less than 10 minutes ×: 10 minutes or more 〇 to ◎, it was considered that the evaluation method was excellent with low complexity.

(2) Detection accuracy The degree of agreement between the quality evaluation result (customer evaluation result) and the quality evaluation result in the film state (in-house evaluation result) after assembling the display device using the inspected retardation film. The detection accuracy was ranked by.

⊚: Evaluation results of 9 to 10 samples match 〇: Evaluation results of 6 to 8 samples match Δ: Evaluation results of 3 to 5 samples match ×: Evaluation results of 0 to 2 samples match Detection accuracy is △ or higher It is required, and it is desirable that it is 〇 to ◎.

Table I shows the above evaluation methods and evaluation results.

From the evaluation results in Table I, by using the alignment unevenness defect detecting method and the alignment unevenness defect detecting device of the retardation film of the present invention, the complexity and detection accuracy of the orientation unevenness defect evaluation of the retardation film can be compared with the visual evaluation. It can be seen that it improves.

Therefore, according to the present invention, by performing quantitative evaluation using an optical system to detect and evaluate orientation unevenness defects in a retardation film, the evaluation is not personal and the evaluation reproducibility can be improved more easily. We were able to provide a method for detecting uneven orientation defects and an apparatus for detecting uneven orientation defects.

The method for detecting uneven orientation defects of a retardation film and the device for detecting uneven orientation defects of a retardation film of the present invention can detect and evaluate uneven orientation defects of a retardation film by quantitative evaluation using an optical system, and the evaluation is personal. Instead, it is suitably used for a simpler evaluation of orientation unevenness defects in a retardation film.

1 Phase difference film 2 Flat LED illumination 3 1st polarizing plate 4 2nd polarizing plate 5 Visual observation 6 Tilt direction L Irradiation light θ Angle 10 Conventional visual evaluation method for alignment unevenness defect 50 Orientation unevenness defect detection device 51 Phase difference film 51a Length 51b Width Hand length 52 Flat LED lighting 53 Shading plate 54 First polarizing plate 55 Second polarizing plate 56 Lens 57 Camera 100 Defect detection unit 101 Control unit 101a CPU
101b RAM
101c Program memory 102 Recording unit 103 Communication unit 104 Data processing unit 104a Shading processing circuit 104b Edge detection circuit 104c Binarization circuit 104d Image processing circuit 104e Filtering circuit 105 Operation display unit 120 Imaging adjustment device 130 Imaging device 150 External output device

Claims

This is a method for detecting uneven orientation defects in a retardation film.
A method for detecting orientation unevenness defects in a retardation film, which comprises the following steps (1) to (5).
Step (1): When the retardation film is arranged between the first polarizing plate and the second polarizing plate arranged on the cross Nicol, the slow axis of the retardation film is the first polarizing plate or the second polarized light. Either the first polarizing plate and the second polarizing plate, or the retardation film so as to be within the range of -10 to 10 ° (excluding 0 °) with respect to the absorption axis of either of the plates. The step of rotating and arranging.
Step (2): A step of irradiating the retardation film with inspection light via the first polarizing plate.
Step (3): A step of photographing the retardation film with a photographing apparatus via the second polarizing plate to obtain a luminance image.
Step (4): The captured luminance image is differentiated (difference) in a direction within the range of 35 to 55 ° or -55 to -35 ° with respect to the slow axis of the retardation film to emphasize the edge portion. Steps to obtain a brightness image.
Step (5): The edge-enhanced luminance image is binarized at a predetermined threshold value to obtain bright pixels above the threshold value and dark pixels below the threshold value, and the pixels are oriented from the bright pixels or dark pixels. Step to detect unevenness.
Further, expansion processing is performed on the binarized luminance image in a diagonal direction within a range of 70 to 120 ° with respect to the direction of the differential (difference) processing of the retardation film, and the diagonal direction is changed. The method for detecting uneven orientation defects in a retardation film according to claim 1, further comprising a step (6) of performing image processing for emphasizing the components.
The method for detecting an orientation unevenness defect of a retardation film according to claim 2, further comprising a step (7) of performing a shrinkage treatment in the same direction with respect to the luminance image after the expansion treatment.
Further, it is provided with the step (8) of extracting the shrinkage-processed luminance image that satisfies the filter condition defined by the representative length of each pixel region as the orientation unevenness defect component in the direction. The method for detecting orientation unevenness defects in a retardation film according to claim 3.
The angle of view θ of the lens calculated from the distance between the lens tip position of the photographing apparatus and the retardation film and the size of the inspection region on the retardation film in the longitudinal direction is the longitudinal direction of the retardation film. The method for detecting an orientation unevenness defect of a retardation film according to any one of claims 1 to 4, wherein the distance is within 23 °.
A device for detecting uneven orientation of a retardation film used in the method for detecting uneven orientation defects of a retardation film according to any one of claims 1 to 5.
The first polarizing plate and the second polarizing plate arranged on the cross Nicol so as to sandwich the retardation film,
A light source that irradiates the retardation film with inspection light via the first polarizing plate,
An imaging device that photographs the retardation film via the second polarizing plate to obtain a luminance image.
A defect detection unit that detects defects from the luminance image is provided.
The defect detection unit
When the retardation film is arranged between the first polarizing plate and the second polarizing plate arranged on the cross Nicol, the slow axis of the retardation film is either the first polarizing plate or the second polarizing plate.

Rotate either the first polarizing plate and the second polarizing plate or the retardation film so as to be within the range of -10 to 10 ° (excluding 0 °) with respect to the absorption axis of. The luminance image taken by arranging the film is differentiated (difference) in a direction within the range of 35 to 55 ° or -55 to -35 ° with respect to the slow axis of the retardation film to emphasize the edge portion. Edge detection circuit and
A binarization circuit that binarizes the edge-detected luminance image with a predetermined threshold value and makes each pixel a bright pixel equal to or higher than the threshold value and a dark pixel lower than the threshold value.
An orientation unevenness detection device for a retardation film, which comprises.
The defect detection unit
Expansion processing is performed on the binarized luminance image in an oblique direction within a range of 70 to 120 ° with respect to the direction of differentiation (difference) processing of the retardation film, and the component in the oblique direction is performed. The orientation unevenness defect detecting apparatus for a retardation film according to claim 6, further comprising an image processing circuit that emphasizes.
The defect detection unit
The orientation unevenness defect detecting apparatus for a retardation film according to claim 7, further comprising an image processing circuit that performs shrinkage processing in the same direction with respect to the luminance image after performing the expansion treatment.
The defect detection unit
6. The feature is that the luminance image subjected to the shrinkage processing has a filtering circuit that extracts a filter condition that satisfies the filter condition defined by the representative length of each pixel region as an orientation unevenness defect component in the direction. The device for detecting uneven orientation of a retardation film according to the above.
The angle of view θ of the lens calculated from the distance between the lens tip position of the photographing apparatus and the retardation film and the size of the inspection region on the retardation film in the longitudinal direction is the longitudinal direction of the retardation film. The uneven orientation defect detecting apparatus for a retardation film according to any one of claims 6 to 9, wherein the distance is 23 ° or less.