JP3507171B2 - Lenticular lens sheet defect inspection method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [0001] BACKGROUND OF THE INVENTION 1. Field of the Invention
Lenticular lens sheet used for screen
Lenticular lens system
In-line white defect detection for black stripes
The present invention relates to an automatic inspection apparatus method when performing the inspection. [0002] 2. Description of the Related Art For transmission type projection screens.
Lenticular lens sheet is made of acrylic resin, polycarbonate
Made of a transparent resin such as carbonate resin, as shown in FIG.
In general, the light source side when used as a screen
Lenticular lens units 510 and 511 are provided on the light exit surface side.
To increase contrast and sharpen the image,
Blacks between the lenticular lenses 510 on the light exit surface side
A light-shielding stripe portion called a tripe 530 is provided.
Has been established. This black stripe is called shading
The lenticular lens sheet 51
0 extruded with acrylic resin, polycarbonate resin, etc.
After shaping, continue to pass between the rolls
The film part is printed and manufactured, and the light-shielding film part is manufactured.
Partly missing or damaged during or after
(Hereinafter referred to as a white defect). Because of this,
When used as an external
May cause loss of contrast and image sharpness.
Identify the location and correct it, or
Must be selected and removed.
It needed to be identified. Conventionally, on a lenticular lens sheet
As a visual inspection method for detecting whitening defects, human visual inspection
In recent years, the transmission type
Lenticular lens sheet for
As a result, high quality and mass production have been demanded.
The conventional visual inspection is different from person to person.
Also, there is a problem with reproducibility.
Need for compact and inexpensive automatic inspection equipment that detects defects
It has come to be. It is also automated for mass production.
Inspection with the used equipment has been required. Corresponding to this
Therefore, the present inventor and the like have shown that
At least, the moving direction of the lenticular lens sheet 740
The lenticular lens sheet in a direction substantially perpendicular to
A predetermined linear region 740 is set as an imaging field of view 750, and reflected night vision is performed.
CCD as a linear area imaging means for imaging with field light
Line sensor camera 710 and reflective darkfield illumination for imaging
A light source that is a lighting means for supplying light720 and, Linear area shooting
Defect detection based on image data obtained by imaging means
And an image processing hand portion 730 for performing
Lenticular lens sheet 740
At a constant speed (in the direction of the black stripe)
While moving, the lenticular lens sheet 740
Defect inspection methods for detecting defects have been proposed. this
In the method, simply the lenticular lens sheet 7
Light from the CCD 40 is sent to the CCD
An imaging field of view from almost directly above the chicular lens sheet 740
Is directly in the direction of movement of the lenticular lens sheet 740.
Shooting with reflective darkfield illumination
Was. However, higher quality and mass production are increasingly required.
In the above method, how much CCD
Even if the field of view of the line sensor camera is small (increase the magnification
Also) Detect white defects in black stripes of 300 μm or less.
I couldn't get it out, which was a problem. 300 μm or less
Can detect white defect of black stripe below
The reason is that there is no lenticular lens inside
S / N of defect detection is reduced due to diffusion material mixed
To improve the S / N ratio.
No suitable method has been proposed. [0004] As described above, the wrench is
Illuminating the reflective lens with dark field illumination
Therefore, image processing is performed on the captured image to
Stripe white defects detected on production line (inline)
And higher quality, mass production
A defect inspection method that can respond to
Came. Under such circumstances, the present invention
-Reflective dark field illumination of the lens
Lenticular lens
Inspection to detect white defects in black stripes
In the method, the S / N of the signal for defect detection is improved.
To provide an automatic inspection method that can respond to increasingly higher quality.
Is to try. [0005] SUMMARY OF THE INVENTION A lenticular of the present invention
Lens sheet defect inspection method should be at least
Or a continuous plate-shaped lenticular lens sheet
One image for imaging a constant straight line area with reflected dark field light
A plurality of linear region imaging means;
On the other hand, an illuminating means for supplying a reflected dark-field illumination for imaging
Based on the image data obtained by the linear region imaging means.
Image processing means for detecting a defect
While moving the lens sheet at a constant speed,
Of the lens lens sheetWhite with black stripe
A defect inspection method for detecting a defect, wherein the lighting means
The angle between the lenticular lens sheet surface and the surface
Set to 1 to image lenticular lens sheetDo
WhenLinear area imaging means for imaging with reflected dark field light and its
By controlling the rotation with the lighting means,The predetermined straight line
Imaging of linear area imaging means for imaging an area with reflected dark field light
The value of the angle θ that the image field makes with the direction of the lenticular lens
Pretend, the image signal obtained by the linear area imaging means
Then, the differential signal S1 of the defect corresponding to various θ values and the defect
The standard deviation N1 of the peripheral differential signal is obtained, and the differential of the defective portion is calculated.
The ratio between the signal S1 and the standard deviation N1 of the differential signal around the defect
Obtain an angle θ2 at which S1 / N1 becomes a maximum, and image a linear area.
The angle between the lenticular lens sheet surface of the means and θ2
It is characterized by setting and imaging. the above,
The standard deviation of the differential signal S1 at the defect and the differential signal around the defect
The difference N1 is used to perform actual defect detection using the second derivative signal.
In this case, it is preferable to use the second derivative signal. still,
As shown in FIG. 3 (a), in the above,
The angle θ1 between the lens surface and the lens
A predetermined straight line area photographed by the imaging means from the step,
The direction to the shadow field area and the lenticular lens sheet surface
Angle between the imaging means and the lenticular lens system.
The direction of the lenticular lens of the board, that is, as shown in FIG.
The angle θ between the direction of the black stripe and the direction of the black stripe
Direction to a predetermined straight line area to be photographed from the camera, that is, the photographing visual field area
And the angle between the lenticular lens direction. [0006] The defect detection of the lenticular lens sheet of the present invention.
The inspection method is as follows:
Reflection dark field illumination of the lens, linear area imaging means
Image processing on the image taken by
Automated inspection methods to detect increasingly higher quality
It enables a possible automatic inspection method. For details,
The positional relationship between the area imaging means and the lighting means is determined by the lighting means.
Of the lenticular lens sheet surface
Is set to a predetermined angle θ1, the predetermined straight line area
Of the linear area imaging means for imaging the area with reflected dark field light
Lenticular lens direction of the chicular lens sheet
(The direction of the black stripe)
Pretend, the differential signal S1 of the defect corresponding to various values of θ and
The standard deviation N1 of the differential signal around the defect is obtained, and
The differential signal S1 and the standard deviation N1 of the differential signal around the defect
Of the angle S2 at which the ratio S1 / N1 of the
The angle between the image means and the lenticular lens sheet surface is θ2
Lenticular lens by setting to and imaging
Defect detection due to the effect of diffusion material mixed inside the sheet
The S / N reduction is reduced as much as possible. [0007] DESCRIPTION OF THE PREFERRED EMBODIMENTS Defects of the lenticular lens sheet of the present invention
Examples of the inspection method will be given, and the present invention will be described below based on the drawings.
explain. Lenticular lens sheet defects of Examples
The inspection method isCalled black stripe 530 shown in FIG.
Devour,Defects in light-shielding film in light-shielding stripes
This is a defect inspection method for detecting (white defects).
The schematic configuration of the detection unit is as shown in FIG.
ing. FIG. 1A is a top view of the defect detection device, and FIG.
FIG. 1B is a diagram viewed from the X direction shown in FIG.
FIG. 2C is a diagram viewed from the Y direction shown in FIG. Figure
2 is an image pickup means (CCD line cell) of the defect detection unit shown in FIG.
FIG. 2 is a diagram showing the relevance of the image processing apparatus and the like. Figure
1. In FIG. 2, 11, 12, 13, and 14 denote CCD line cells.
Sensors, 21, 22, 23 and 24 are fluorescent lamps, 11
a, 12a, 13a and 14a are CCD line sensor cameras
La controller, 30 is a lenticular lens sheet,
Reference numerals 41, 41a, 42b, and 42 denote imaging visual fields, and 50 denotes an image processing.
1 shows a processing device. As shown in FIG. 1, for defect detection
The inspection device of the lenticular lens sheet 30 passing surface
Above the two CCD line sensor cameras 11 and 12
And imaging means 21 and 22 are provided.
Image of the imaging visual field 41 with reflection dark field illumination, and two C
Imaging means including CD line sensor cameras 13 and 14;
Illumination means consisting of 23 and 24 is provided for reflection darkfield illumination
The imaging visual field 42 is imaged by
In the direction of conveyance of the lenticular lens sheet.
CCD line sensor camera 1 so as to straddle the intersection width
1, 12, 13, and 14 are arranged. With this device configuration
The lenticular lens sheet 30 is as shown in FIG.
The direction of the lenticular lens, that is, black strike
CCD line sensor camera 1 while moving in the direction of
Images are taken at 1, 12, 13, and 14. In FIG. 1 (c), C
It comes out of the CD line sensor cameras 11 and 12, respectively.
The two lines represent the field of view taken by the camera and the CCD
The in-sensor camera 11 is a lenticular lens sheet 3
The imaging field of view 41a, which is a linear area on 0, is
Sensor camera 12 on lenticular lens sheet 30
Of the imaging field of view 41b, which is a linear area of
The field 41 is imaged. Although not shown, the CCD
The in-sensor cameras 13 and 14 each have an imaging field of view 42.
a and 42b are imaged, and the imaging visual field 42 is also imaged.
CCD line sensor camera controller 1 shown in FIG.
1a, 12a, 13a, and 14a are CCD lines, respectively.
Field of view 41 of the non-sensor cameras 11, 12, 13, 14
a, 41b, 42a, 42b are lenticular lenses
The angle between the lenticular lens of the
To control the value of the angle θ formed with the stripe direction.
When the fluorescent light 21 and 22
The angle between the lens and the surface of the
Lights 23 and 24 are lenticular in imaging field of view 42
The angle between the lens sheet 30 and the surface is set to a predetermined angle θ1.
CCD line sensor cameras 11, 12, 1
3, 14 and fluorescent lamps 21, 22, 23, 24 were paired.
The rotation is controlled in the state. In addition, CCD line cell
The direction in which the sensor cameras 11, 12, 13, and 14 capture images is
The direction perpendicular to the surface of the lenticular lens sheet 30
It is a direction. The inspection method of this embodiment is described in FIGS.
Defect detection using a device for defect detection shown below
Through the lenticular lens sheet 30 at a constant speed.
When passing, two CCD line sensor cameras 1
Imaging of lenticular lens sheet 30 by 1 and 12
Takes an image of the visual field area 41 and uses two CCD line sensor
Lenticular lens sheet 30
The imaging of the imaging visual field area 42 is performed.
Thus, the lighting means 21, 22, 23, 24 are
Angle θ1 with respect to the lens 30
In this case, the light from each illumination means is
Fields 41 and 42 are imaged.
Lenticular lens of lenticular lens sheet 30
Angle, that is, the angle θ with the direction of the black stripe
Was adjusted so that the S / N of defect detection was as large as possible.
imaging for defect detection at an angle θ2 of θ.
This improves the quality of the lenticular lens 30
It can respond to. CCD line sensor turtle
Image signals (data) obtained by the cameras 11, 12, 13, and 14.
2) are processed by the image processing unit 50 shown in FIG.
Lenticular lens sheet 30
The image signals of all predetermined regions are processed.
The method is as follows: CCD line sensor cameras 11, 12, 13,
14 can provide an image signal (data) S / N obtained
Just trying to make it bigger. In the inspection method of this embodiment,
Of the lenticular regular lens sheet 30
The angle between the lens and the lens
Angle θ2 suitable for defect detection
I asked. A specific procedure will be described with reference to FIG. Destination
First, as shown in FIG.
0 is the place where the illumination means is the lenticular lens sheet 310
Light irradiation to imaging field of view area 340 with certain defects (white defects)
Direction and the angle between the lenticular lens sheet surface
A predetermined angle θ1 (here, θ1 is 30 °)
Set as follows. Next, the CCD line sensor camera 3
10 imaging fields of view 340 and lenticular lens sheet 3
Angle with 30 lenticular lenses, ie black
While changing the value of the angle θ formed with the direction of the stripe,
The second derivative signal S1 of the defect corresponding to each θ value and the defect
A peripheral second derivative signal standard deviation N1 is detected. Differentiation
The reason is to emphasize the signal of the defective part,
Noise that cannot be removed is also detected. Then each obtained
The second derivative signal S1 of the defect corresponding to the value of θ and the periphery of the defect
Find the ratio S1 / N1 of the standard deviation N1 of the second derivative signal of the side
You. When the ratio S1 / N1 is graphed, as shown in FIG.
And the value of S1 / N1 is such that the value of θ is greater than 60 °.
When the angle is between 60 ° and 0 °,
It can be seen that the same value is shown. Therefore, the defective part
Assuming that the second derivative signal S1 is a signal S, the second derivative around the defect is
The standard deviation N1 of the differential signal is approximately equal to the maximum value of the noise signal N.
Since it is proportional, the S / S
The angle θ2 that increases N is the ratio S in dark-field reflection imaging.
1 / N1 should be less than 60 ° which is a large value range
It turns out that it is good. Thus, the lighting means 320
Predetermined angle to the lenticular lens sheet surface
CCD line corresponding to the angle of 30 °
Field of view of lenticular camera 310 and lenticular lens
Angle θ2 between the lens sheet 330 and the lenticular lens
Is determined in a range of about 60 ° or less. In this manner, the angle θ suitable for detecting a defect is
The reason why it is necessary to determine is based on the following reasons. wrench
A diffusing material is dispersed inside the resin of the lens sheet.
And this is in a random direction,
Due to the influence, the CCD line sensor camera 60 is
In the same place as the lenticular lens sheet 30
A signal that changes more locally will be detected. others
In the same way as the differential processing in the defect inspection described later,
It cannot be removed even by minute processing, and as a result,
It remains as a noise signal in addition to the defective signal. This is
This is the first reason. And lenticular lens sheet
Diffusion material is random inside resin
But it is incident on the lenticular lens sheet
Depending on the direction of the light emitted from the diffuser
The noise signal (light intensity) changes differently than the defect signal changes
Is the second reason. Thus, the fluorescent lamp as the illumination means
21, 22, 23, 24 and a CCD line as an imaging means.
Of the field of view of the sensor cameras 11, 12, 13, and 14
After the direction (angle θ) is determined, actual defect detection is performed.
However, CCD line sensor camera 1 for defect detection
The processing of the image data from 1, 12, 13, and 14 is shown in FIG.
Is performed by the image processing apparatus 50 shown in FIG. In addition, as shown in FIGS.
In the apparatus, the image processing unit 50 determines an image when determining the angle θ.
CCD line sensor camera 1 that also serves as an image processing unit
1, 12, 13, 14 imaging field of view and lenticular lens
Controller that adjusts the angle θ formed by the angle. Picture
When detecting a defect, the image processing unit 50 controls the CCD line sensor.
Image signals obtained from cameras 11, 12, 13, and 14
(Data), first-order differential processing using a spatial filter
Or perform a second-order differentiation process, and on the obtained data,
Further, slice processing is performed to exceed a predetermined threshold.
An object is detected as a defect. The lenticular used in this embodiment
-The material of the lens sheet is acrylic (refractive index 1.51)
The diffusing material is glass having a refractive index of 1.53 to 1.54,
With an average size (particle size) of 17 μm (MAX 38 μm)
And 4% by weight in acrylic. This embodiment
The defect detection method of the above
Lack of imaging field of view of imaging means (CCD line sensor camera)
It determines the appropriate angular position for detecting the defect,
It is possible to improve the S / N for detecting a white defect in a black stripe portion.
Things. In the case of the conventional defect detection unit configuration shown in FIG.
Could detect only defects up to 300 μm square
However, in this embodiment, θ is set to 60 degrees or less.
With this, up to a defect corresponding to a 200 μm square area,
It is now possible to stably detect only defects
Was. Next, as described above, the lighting means
Fluorescent lamps 21, 22, 23, 24,
Position of CCD line sensor camera 11, 12, 13, 14
Image data obtained by imaging the device with the
A process for detecting a defect in the image processing unit 50 shown in FIG.
And explain briefly. Image processing unit 50 shown in FIG.
Is the first derivative processing or the image signal obtained by imaging
By performing second derivative processing, signal changes are obtained and defects are detected.
To specify, the differentiation process here is an element array
Using a filter with
The calculation is performed. In addition, here, from the imaging means
The obtained image data P (X_ij) And space filter
Bull W (i, j) and P (X_ij) And W (i, j)
Performing the multiply-accumulate operation is called filtering.
Filter W (i, j) used in the processing of
In general, a differential filter or a spatial filter
Call. Differential file for Y direction array of pixel data
Is the sum of the pixel of interest and the pixels before and after it.
And the direction of the array, as shown in FIG.
You. For example, the image data P (Y_n), While FIG.
(+1, -2, +1) and space filter table
(Y table)
In this way, a second differentiation process in which smoothing is performed for two pixels can be performed. Attention
Pixel is Y_n, The image data P (Y_n) And space
Sum of product operation S with the ilter table_nIs (+1) × Y
_n-1+ (-2) × Y_n+ (+ 1) × Y_{n + 1}Becomes Ma
Also, as shown in FIG. 6B, the spatial filter is added to (-1, +1).
-By setting the table (weight table)
Differentiation can be performed. The image processing unit 50 is a CCD detector of the defect detection unit.
Each image from the in-sensor cameras 11, 12, 13, and 14
For data (signal), for example,
Detect a fall. Fig. 7 shows one CCD line sensor camera
Image processing for image data (signal) obtained from
FIG. Second derivative in FIG.
Is based on sampling by a CCD line sensor camera.
Line data obtained from
The spatial filter table and the
The product-sum operation of
Store the line data obtained by the ring
Is performed between the line data items. In the case of FIG. 7, 9
This process is performed between line data, but a new line data
Each time the data is sampled, the older line data is
This process is always performed with 9 line data.
In FIG. 7A, D_ijIn the j-th line data
represents the pixel data at the i-th position,
When the sum of the pixel data in the product is calculated between the lines,
The results of the differentiation process and the primary differentiation process are shown in FIG.
The result is as shown in FIG. Obtained in this way
In addition, each operation result of the second derivative processing or each
The result of the operation is further sliced to a predetermined value.
Detect defects that exceed the threshold. Here,
Line data refers to the image pickup means (CCD line sensor camera)
B) Transfer of lenticular lens sheet
A straight line on the sheet surface that straddles the width in the direction substantially perpendicular to the movement direction
Imaging of a field of view including a predetermined width on a line in one sampling
In the data, say the data corresponding to the whole area of the specified width
I have. In order to implement the defect inspection method of the present invention,
In addition to the device configuration shown in FIG.
The structure shown in FIG. FIG. 4 (a)
FIG. 4B is a top view of the defect detection device, and FIG.
FIG. 4C is a view from the X direction shown in FIG.
FIG. 3 is a diagram viewed from a Y direction. In the case of the device configuration shown in FIG.
Are CCD line sensor cameras 11A, 12A, 13
A, 14A are the respective imaging visual fields 41Aa, 41Ab,
42Aa, 42Ab (42Aa, 42Ab
No), and if the defect has directionality
It is effective in the case. Note that the tilt shooting is as shown in FIG.
Around the angle of view of the imaging system (lens system) when imaging the sample
It is taken by the department. The device configuration of the defect detection unit shown in FIG.
When the bright defect inspection method is implemented with θ2 = 0 °
It is. FIG. 5A is a top view of the defect detection device, and FIG.
FIG. 5B is a diagram viewed from the X direction shown in FIG.
FIG. 5C is a diagram viewed from the Y direction shown in FIG. Figure
In the case of the device configuration shown in FIG. 5, a CCD line sensor camera
11B and 12B are imaging fields of view 41Aa and 41A, respectively.
Image b. In the case of the device shown in FIG.
-It is possible to detect a defect in the moving direction of the lens sheet 30B.
Thus, it is moving in a different direction from FIG. [0017] According to the present invention, as described above, a lenticular
-Reflective dark field illumination of the lens sheet, linear area imaging means
Image processing on the image taken by
Higher quality, more quantity in automatic inspection equipment to detect
Lenticular lens sheet black
To provide a method for detecting white defects in mask stripes
You. Further, the present invention provides various types of lasers having different shapes, dimensions, and materials.
Defect signal for defect inspection of lens lens sheet
Can be improved.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a schematic view of an apparatus for implementing a lenticular lens sheet defect detection method of the present invention. FIG. 2 is an apparatus for implementing a lenticular lens sheet defect detection method of the present invention. FIG. 3 is a diagram for explaining a method for detecting a defect of a lenticular lens sheet according to an embodiment. FIG. 4 is a schematic diagram of an apparatus for performing a method for detecting a defect on a lenticular lens sheet according to the present invention. FIG. 6 is a schematic view of an apparatus for implementing the lenticular lens sheet defect detection method according to the first embodiment. FIG. 6 is a diagram illustrating a spatial filter. FIG. 7 is a diagram illustrating image processing in the embodiment. FIG. FIG. 9 is a view showing a defect of a lenticular lens sheet. FIG. 10 is a schematic view of an apparatus for explaining a conventional defect inspection method. , 13, 14 CCD line sensor cameras 21, 22, 23, 24 Fluorescent lamps 11a, 12a, 13a, 14a CCD line sensor camera controller 30 Lenticular lens sheets 41, 41a, 42b Imaging fields 42, 42a, 42b Imaging fields 50 Image processing Device 310 CCD line sensor camera 320 Fluorescent lamp 330 Lenticular lens sheet 340 Image field of view 11A, 12A, 13A, 14A CCD line sensor camera 30A Lenticular lens sheet 41A, 41Aa, 42Ab Image field of view 42A, 42Aa, 42Ab Image field 11B, 12B, 13B , 14B CCD line sensor camera controller 30B Lenticular lens sheet 41B, 41Ba, 42Bb Imaging field of view 42B, 42Ba, 42Bb Imaging field of view 510 wrench Interview Ra lens sheet 511 lenticular lens portion 530 black stripes 710 CCD line sensor camera 720 light source 730 image processor 740 lenticular lens sheet 750 imaging field

Continuation of the front page (56) References JP-A-3-15067 (JP, A) JP-A-63-295948 (JP, A) JP-A-1-260585 (JP, A) JP-A-2-116741 (JP) JP-A-6-258249 (JP, A) JP-A-1-169343 (JP, A) JP-A-3-130545 (JP, U) (58) Fields investigated (Int. Cl. ⁷ , DB G01M 11/00-11/02 G01B 11/00-11/30 102 G01N 21/84-21/958 G03B 21/60 G06T 1/00-1/40 G06T 3/00-3/60 G06T 5 / 00-5/50 G06T 7/00-7/60 G06T 9/00-9/40

Claims (1)

(57) [Claim 1] At least one or a plurality of lenses for imaging a predetermined linear region of a lenticular lens sheet having a sheet-like plate shape or a continuous plate-like shape with reflected dark field light. A linear region imaging unit, an illumination unit that supplies reflected dark field illumination for imaging to the linear region imaging unit, and an image for detecting a defect based on image data obtained by the linear region imaging unit Processing means for detecting a white defect in a black stripe portion of the lenticular lens sheet while moving the lenticular lens sheet at a constant speed, wherein the illuminating means forms an angle with the lenticular lens sheet surface. Is set to a predetermined angle θ1, when imaging the lenticular lens sheet, the reflected dark field light
The linear area imaging means for imaging and its illumination means are paired.
Then, by controlling the rotation, the value of the angle θ formed by the imaging field of view of the linear area imaging means for imaging the predetermined linear area with the reflected dark field light is obtained by the linear area imaging means. For the obtained image signal, a standard deviation N1 of the differential signal S1 of the defect and a differential signal of the differential signal of the peripheral portion of the defect corresponding to various θ values is obtained, and the standard deviation S1 of the differential signal of the defective portion and the standard deviation of the differential signal of the peripheral portion of the defect are obtained. A defect inspection method for a lenticular lens sheet, characterized in that an angle θ2 at which a ratio S1 / N1 with respect to N1 is maximized is determined, an angle between the linear area imaging means and a lenticular lens sheet surface is set to θ2, and imaging is performed.