JP3357968B2 - Lenticular lens sheet defect inspection method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for inspecting the appearance of a lenticular lens sheet used for a transmission type projection screen, and more particularly to an automatic method for detecting in-line white defects of a black stripe portion of a lenticular lens sheet. The present invention relates to an inspection apparatus method.

[0002]

2. Description of the Related Art A lenticular lens sheet for a transmission type projection screen is made of a transparent resin such as an acrylic resin or a polycarbonate resin. As shown in FIG. 6, generally, as shown in FIG. Lenticular lenses 61 and 62 are provided on the sides, and a light-shielding stripe portion called a black stripe 63 is provided between the lenticular lenses 61 on the light exit surface side to increase the contrast and sharpen the image. The light-shielding stripes called black stripes are produced by extruding the lenticular lens sheet 60 with an acrylic resin, a polycarbonate resin, or the like, and then printing the light-shielding film while continuously passing between rolls. , In the process of producing the light-shielding film portion, or in a portion where it is partially missing or damaged thereafter (hereinafter,
White defects). For this reason, when used as a screen, reflected light from the outside is a cause of impairing contrast and image sharpness,
It was necessary to identify and correct the location or to selectively remove the screen, and it was necessary to identify the location of the white defect.

Conventionally, as a visual inspection method for detecting the above-mentioned white defect of a lenticular lens sheet, a visual inspection method has been adopted. However, recently, a lenticular lens sheet for a transmission type projection screen has been used in a high degree. The demand for quality and mass production is increasing, and conventional inspection with the naked eye has differences between people and has problems with reproducibility. Inexpensive automatic inspection equipment has been required. In addition, an inspection using an automated device has been required for mass production. In order to cope with this, the inventors of the present invention have made the lenticular lens sheet 74 at least substantially perpendicular to the moving direction of the lenticular lens sheet 74 as shown in FIG.
A CCD line sensor camera 71 which is a linear area image pickup means for taking an image with reflected dark field light using a predetermined straight line area as an imaging field of view 75, and a light source which is a lighting means for supplying reflection dark field illumination for image pickup 72 and an image processing unit 73 for detecting a defect based on image data obtained by the linear region imaging means, and the continuous plate-shaped lenticular lens sheet 74 is moved in the direction of the lenticular lens (black stripe direction). Moving at a constant speed)
A defect inspection method for detecting a defect of the lenticular lens sheet 74 has been proposed. In this method, the light from the lenticular lens sheet 74 is simply
By the line sensor 71, almost directly above the lenticular lens sheet 74, that is, the lenticular lens sheet 74
The image was taken by the reflection dark field illumination so that the angle formed by the surface was approximately 90 degrees. However, higher quality and mass production are increasingly required. In the above-mentioned method, even if the field of view of the CCD line sensor camera is small (even if the magnification is increased), the white of the black stripe of 300 μm or less can be obtained. The defect could not be detected, which was a problem. The reason why the white defect of the black stripe of 300 μm or less cannot be detected is that the diffusion material mixed in the lenticular lens sheet causes
It has been said that the S / N for defect detection is reduced, but no suitable method for improving the S / N has been proposed.

[0004]

As described above, the lenticular lens is illuminated by reflection in a dark field, image processing is performed on an image photographed by the linear region imaging means, and a white defect of a black stripe of a sample is produced on a production line (in-line). As for the automatic inspection method for detecting defects, a defect inspection method capable of coping with higher quality and mass production has been required. Under such circumstances, the present invention provides an automatic inspection for illuminating a lenticular lens with reflection in a dark field, performing image processing on an image captured by a linear area imaging unit, and detecting a white defect in a black stripe portion of the lenticular lens. It is an object of the present invention to provide an automatic inspection method capable of improving the S / N of a signal for defect detection and responding to higher quality.

[0005]

A method of inspecting a lenticular lens sheet for defects according to the present invention comprises the steps of: (a) extending at least a width in a direction substantially perpendicular to a moving direction of a sheet-shaped or continuous plate-shaped lenticular lens sheet; One or more linear region imaging means for imaging a predetermined linear area of the lenticular lens sheet with reflected dark field light, and reflection dark field illumination for imaging the linear area imaging means. The lenticular lens sheet is provided with an illuminating means for supplying and an image processing means for detecting a defect based on the image data obtained by the linear region imaging means. In the defect inspection method for detecting a lenticular lens, the angle formed by the lighting means with the lenticular lens sheet surface is set to a predetermined angle θ1, -When the lens sheet is imaged, the value of the angle θ formed with the lenticular lens sheet surface of the linear area imaging means for imaging the predetermined linear area with the reflected dark field light is determined, and defective portions corresponding to various θ values are detected. Differential signal S1
And the standard deviation N1 of the differential signal of the defect peripheral portion are obtained, and the differential signal S1 of the defective portion and the standard deviation N1 of the differential signal of the defect peripheral portion are obtained.
The angle θ2 at which the ratio S1 / N1 with respect to the maximum is obtained is determined, the angle between the linear region imaging means and the lenticular lens sheet surface is set to θ2, and imaging is performed.
When the actual defect detection is performed by using the secondary differential signal, the secondary differential signal is preferably used as the standard deviation N1 between the differential signal S1 of the defective portion and the differential signal of the peripheral portion of the defect. In addition, as shown in FIG. 3A, in the above description, the angle θ1 formed by the illuminating unit with the lenticular lens sheet surface is defined as the direction from the illuminating unit to a predetermined linear region photographed by the imaging unit, that is, the direction to the photographing visual field region. The angle θ between the lenticular lens sheet surface and the imaging means and the lenticular lens sheet surface is the angle θ between the direction to a predetermined linear area photographed by the imaging means, that is, the direction to the photographing visual field area, and the lenticular lens sheet surface. It is an angle, and the illumination means and the imaging means are opposite to each other with respect to the normal in the field of view of the lenticular lens surface.

[0006]

According to the defect inspection method for a lenticular lens sheet of the present invention, by adopting the above-described structure, the lenticular lens is illuminated by reflection in a dark field and subjected to image processing on an image photographed by the linear region imaging means. In the automatic inspection method for detecting a defect, an automatic inspection method capable of responding to increasingly higher quality is enabled. More specifically, the positional relationship between the linear region imaging unit and the illuminating unit reflects the predetermined linear region when the angle between the illuminating unit and the lenticular lens sheet surface is set to a predetermined angle θ1. The value of the angle θ formed with respect to the lenticular lens sheet surface of the linear area imaging means for imaging with dark field light is changed, and the differential signal S of the defective portion and the differential signal of the peripheral portion of the defect corresponding to various values of θ are obtained. The standard deviation N is obtained, and the ratio S / N between the differential signal S of the defective portion and the standard deviation N of the differential signal at the peripheral portion of the defect is obtained.
Is obtained, the angle between the lenticular lens sheet surface of the linear region imaging means and the angle formed with the lenticular lens sheet surface is set to θ2, and imaging is performed to detect a defect caused by the diffusion material mixed in the lenticular lens sheet. The S / N reduction is minimized.

[0007]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of a method for inspecting defects of a lenticular lens sheet according to the present invention will be described below with reference to the accompanying drawings. The defect inspection method of the lenticular lens sheet of the embodiment is a defect inspection method for detecting a defect (white defect) of a light-shielding film portion of a light-shielding stripe portion called a black stripe 630 shown in FIG. FIG. 1 schematically shows the entire functional configuration of the apparatus for detecting a defect. FIG. 2 is a diagram showing the relevance of the imaging means (CCD line sensor), image processing device, and the like of the defect detection device shown in FIG. 1 and 2, reference numeral 10 denotes illumination means, 20A and 20B denote CCD line sensor cameras, 21
A and 21B are CCD line sensor camera controllers,
30 is a lenticular lens sheet, 40, 40A, 4
0B denotes an imaging field of view, and 50 denotes an image processing device. As shown in FIG. 1, the apparatus for detecting a defect is provided with an imaging unit 20 including two CCD line sensor cameras 20A and 20B and an illuminating unit 10 above the lenticular lens sheet 30 passing surface. In FIG. 1B, two lines respectively protruding from the CCD line sensor cameras 20A and 20B represent the field of view taken by the cameras. The imaging unit controller 21 shown in FIG. 2 is for adjusting the value of the angle θ formed by the imaging unit 20 with respect to the lenticular lens sheet 30 surface.

In the inspection method of the present embodiment, the defect is detected by using the defect detecting device shown in FIG. 1 and FIG. 2, and the lenticular lens sheet 30 is passed at a constant speed. 2 CCD line sensor cameras 2
0A and 20B respectively image the imaging visual field regions 40A and 40B of the lenticular lens sheet 30 to obtain an entire imaging visual field 40 of a predetermined area of the lenticular lens sheet 30. Is set at a predetermined angle θ1 with respect to the lenticular lens 30, the lenticular lens 3 of the CCD line sensor cameras 20A and 20B that images the imaging visual field 40 using the light from the illumination means 10 as dark-field illumination.
The angle θ with respect to 0 is optimized, and thereby, it is possible to cope with the high quality of the lenticular lens 30.

[0009] CCD line sensor cameras 20A, 20B
As shown in FIG. 2, the image signals (data) obtained by the above are processed by the image processing unit 50, and the image signals of the entire imaging visual field 40 in a predetermined area of the lenticular lens sheet 30 are processed together. You. Image processing is CCD
The primary differential processing or the secondary differential processing is performed using a spatial filter on the image data (signal) obtained by the line sensor cameras 20A and 20B.

Hereinafter, the inspection method of this embodiment is performed as described above, and a specific procedure will be described with reference to FIG.
First, as shown in FIG. 3A, the illuminating unit 10 is formed by a direction in which the illuminating unit irradiates light to the imaging visual field region 40 having a predetermined defect (white defect) of the lenticular lens sheet 30 and the lenticular lens sheet surface. Angle is a predetermined angle 30
Set to be °. Next, the CCD line sensor cameras 20A and 20B and the imaging visual field area 40A, respectively,
While changing the value of the angle θ between the direction of the camera connecting the camera 40B and the lenticular lens sheet 30 as shown in Table 1, the second derivative signal S1 of the defect corresponding to various θ values and the two signals around the defect are obtained. The second derivative signal standard deviation N1 is detected.
The differentiation processing is for emphasizing the signal of the defective portion, but also detects noise that cannot be removed. Then, FIG.
As shown in (b), the ratio S1 / N1 of the standard deviation N1 of the secondary differential signal S1 of the defective portion and the secondary differential signal around the defective portion corresponding to each obtained value of θ is obtained. The ratio S in FIG.
FIG. 3C shows a graph of 1 / N1.
The value of θ indicating the maximum value of 1 / N1 is about 50 °. Here, assuming that the second derivative signal S1 of the defective portion is a signal S, the standard deviation N1 of the second derivative signal around the defective portion is the noise signal N
Is approximately proportional to the maximum value of S / N.
In order to obtain the optimum angle, the angle θ2 that maximizes the ratio S1 / N1 in the dark-field reflection imaging is obtained, and this is calculated as S / N
Can be set to the optimum angle. In this manner, the angle θ between the lenticular lens sheet surface of the linear region imaging unit and the linear region imaging unit suitable for defect detection, which corresponds to the angle formed by the illumination unit 10 with respect to the lenticular lens sheet surface, corresponds to the predetermined angle of 30 °. Determine to be about 50 °.

The reason for determining the angle .theta. Of the linear region imaging means with respect to the lenticular lens sheet surface suitable for defect detection in this manner is as follows. A diffusing material is dispersed inside the resin of the lenticular lens sheet and is oriented in a random direction. Due to the influence of the diffusing material, the CCD line sensor cameras 20A and 20B are exposed to the lenticular lens sheet in the same manner as the defective portion. 3
A signal that locally changes depending on the location of 0 is detected. Therefore, even if the differential processing is performed in the same manner as the differential processing in the defect inspection to be described later, the signal cannot be removed, and as a result, it remains as a noise signal in addition to the defect signal. This is the first reason. The diffusing material is oriented in a random direction inside the resin of the lenticular lens sheet. However, as shown in FIG. 3 (b), depending on the angle of light incident on the lenticular lens sheet,
The second reason is that the noise signal (light intensity) changes differently from the change in the defect signal depending on the direction of the light emitted from the diffusion material.

These processes are performed by the image processing apparatus 5 shown in FIG.
0, and thus, the lighting means 1
0, CCD line sensor camera 20 as imaging means
The actual defect detection is performed after the directions (angle θ) of A and 20B are determined. However, the image processing unit for this defect detection also serves as the image processing when determining the angle θ. The image processing apparatus 50 used in the lenticular lens sheet defect inspection method according to the present embodiment is simple. At the time of actual defect detection, an image signal (data) obtained by the CCD line sensor cameras 20A and 20B is subjected to a first-order differentiation process or a second-order differentiation process by an image processing unit 50 using a spatial filter for the image data (signal). Differentiation processing is performed, and the obtained result is further subjected to slicing processing to detect a defect exceeding a predetermined threshold value as a defect.

The material of the lenticular lens sheet used in this embodiment is acrylic (refractive index 1.51).
The diffusing material is glass having a refractive index of 1.53 to 1.54, an average size (particle size) of 17 μm (MAX 38 μm), and 4% by weight in acrylic. The defect detection method according to the present embodiment determines an optimum angular position for imaging by the linear area imaging means (CCD line sensor camera) by the above-described procedure. Basically, a wrench as shown in FIG. In the case of the cue lens sheet, the white defect detection S /
N can be improved.

Next, as described above, the illumination means 1
0, a process of detecting a defect in the image processing unit 50 shown in FIG. 2 with respect to image data obtained by imaging of the apparatus in which the position of the imaging unit is set will be briefly described. The image processing unit 50 shown in FIG. 2 performs a first-order differentiation process or a second-order differentiation process on an image signal obtained by imaging to obtain a change in a signal and identify a defect. Is performed by performing a product-sum operation for each sampling using a filter having an element array. Here, the image data P (X _ij ) and the spatial filter table W (i, j) obtained from the imaging means are used to calculate P (X _ij ).
Performing the product-sum operation of the data and W (i, j) is called a filtering process, and a table W (i,
j) is called a filter, and is generally called a differential filter or a spatial filter. The differential filter for the Y-direction array of pixel data uses the weight of the pixel of interest and the pixels before and after it in the product-sum operation and the direction of the array, as shown in FIG.
It is represented as For example, the image data P (Y _n), FIG. 4 as in (a) (+ 1, -2 , + 1) and the two pixels smoothing the Y direction by setting the spatial filter table (weight table) The second derivative processing can be performed. If the target pixel was Y _n, the image data P
Product-sum operation S of (Y _n ) and spatial filter table
_n is (+1) × Y _n-1 + (− 2) × Y _n + (+ 1) ×
Yn _{+ 1} . Also, as shown in FIG. 4B, (-1, +
By setting a spatial filter table (weight table) in 1), primary differentiation processing can be performed.

[0015] The image processing unit 50 is provided with the CC of the photographing apparatus 110.
For example, the following processing is performed on each image data (signal) of the D-line sensor cameras 20A and 20B to detect a defect. FIG. 5 is a diagram for explaining image processing on image data (signal) obtained from one CCD line sensor camera. The second derivative in FIG. 5 performs a product-sum operation with a spatial filter table for each pixel data at the same position between lines with respect to line data obtained by sampling with a CCD line sensor camera. Line data obtained by continuous sampling of numbers is accumulated, and a product-sum operation is performed between the line data. In the case of FIG. 5, this processing is performed between the nine line data. However, every time new line data is sampled, this processing is always performed on the nine line data except for the line data that is gradually older. ) Indicates D _ij represents the pixel data at the i-th position in the j-th line data, and when the pixel data at the M-th position is multiply-accumulated between the lines, the results of the secondary differentiation processing and the primary differentiation processing are respectively 5 (b) and FIG. 5 (c). For each operation result of the secondary differentiation process or each operation result of the primary differentiation process obtained in this way,
Further, slice processing is performed to detect a defect exceeding a predetermined threshold value. Here, the line data is
The image data obtained in one sampling of a field of view including a predetermined width on a straight line on the sheet surface across a width in a direction substantially orthogonal to the moving direction of the lenticular lens sheet obtained by the imaging means (CCD line sensor camera). It refers to data corresponding to the entire width.

[0016]

As described above, the present invention relates to an automatic inspection apparatus for detecting a defect of a sample by illuminating a lenticular lens sheet with reflection in a dark field and performing image processing on an image photographed by a linear region imaging means. It is possible to provide a method for detecting a white defect of a black stripe of a lenticular lens sheet, which can cope with higher quality and mass production. Further, the present invention makes it possible to improve the S / N of a defect signal for defect inspection of various lenticular lens sheets having different shapes and dimensions and materials.

[Brief description of the drawings]

FIG. 1 is a schematic view of an apparatus for implementing a method for detecting a defect of a lenticular lens sheet according to an embodiment.

FIG. 2 is an apparatus diagram for implementing a method for detecting a defect of a lenticular lens sheet according to an embodiment.

FIG. 3 is a diagram illustrating a method for detecting a defect of a lenticular lens sheet according to an embodiment.

FIG. 4 is a diagram showing a spatial filter;

FIG. 5 is a diagram for explaining image processing in the embodiment.

FIG. 6 is a view showing a defect of a lenticular lens sheet.

FIG. 7 is a schematic view of an apparatus for explaining a conventional defect inspection method.

[Explanation of symbols]

Reference Signs List 10 illumination means (light source) 20 imaging means 20A, 20B CCD line sensor camera 21A, 21B CCD line sensor camera controller 30 lenticular lens sheet 40, 40A, 40B imaging field of view 50 image processing unit 60 lenticular lens sheet 61, 61 lenticular lens unit 63 Black stripe 71 CCD line sensor camera 72 Light source 73 Image processing unit 74 Lenticular lens sheet 75 Imaging field of view

──────────────────────────────────────────────────続 き Continuation of the front page (56) References JP-A-1-260585 (JP, A) JP-A-1-237441 (JP, A) JP-A-6-82386 (JP, A) JP-A-6-82386 285249 (JP, A) (58) Fields investigated (Int. Cl. ⁷ , DB name) G01N 21/892-21/956 G01M 11/00 G06T 1/00 420

Claims (1)

(57) [Claims] At least a predetermined linear region of a lenticular lens sheet, which straddles a width in a direction substantially perpendicular to the moving direction of a lenticular lens sheet having a single-plate shape or a continuous plate shape, is reflected by reflected dark-field light. One or a plurality of linear area imaging means for imaging, illumination means for supplying reflection dark field illumination for imaging to the linear area imaging means, and an image obtained by the linear area imaging means An image processing means for detecting a defect based on the data, wherein the illuminating means detects a defect in the lenticular lens sheet while moving the lenticular lens sheet at a constant speed. When the lenticular lens sheet is imaged by setting the angle to be a predetermined angle θ1, the predetermined linear region is reflected by dark field light. The value of the angle θ between the lenticular lens sheet surface of the linear area imaging means to be imaged and the standard deviation N1 of the differential signal S1 of the defect portion and the differential signal of the peripheral portion of the defect corresponding to various θ values are obtained. Ratio S1 / N1 of the differential signal S1 of the above and the standard deviation N1 of the differential signal at the periphery of the defect
A method for inspecting a defect of a lenticular lens sheet, comprising: obtaining an angle θ2 at which a maximum value is obtained, setting an angle between the lenticular lens sheet surface of the linear region imaging means to θ2, and imaging.