TW201042295A - Antiglare processing method, method for manufacturing antiglare films and method for manufacturing molds - Google Patents

Abstract

The present invention provides an antiglare processing method for transparent substrates, a method for manufacturing antiglare films and a method for manufacturing molds appropriate for implementing the method. The antiglare processing method includes a step of using a filter to at least remove or reduce low space frequency components from space frequency components contained in a first pattern to generate a second pattern, and a step of producing a convex-concave shape on a transparent substrate based on the pattern. The method may further include a step of generating a third pattern transformed to information which has been discretized by a dithering method, and a step of moving isolated pixels by a Monte Carlo method to generate a fourth pattern.

Description

[Technical Field] The present invention relates to an anti-glare treatment method, a method for producing an anti-glare film, and a method for producing an anti-glare treatment method and an anti-glare film. Production method. Town [Prior Art] Image line device in liquid aa display, plasma display panel, image tube (cathode tube: CRT) display, organic electroluminescence (EL) display, etc., when external light is reflected on its display surface It will significantly impair the recognition. Therefore, in order to prevent the reflection of such external light, the image or the computer, the outdoor camera or digital camera that uses strong external light, and the use of reflected light In the mobile phone or the like for display, a process for preventing external light from being reflected on the face of the video display device is applied. The processing applied to the surface of the image display device can be roughly divided into: non-reflective treatment using interference of the optical multilayer film; and fine scattering at the table (4) to scatter the incident light to dilute the image of the image. Dizzle treatment. The former does not require the formation of a multilayer film having a uniform optical film thickness, so the cost becomes high. In contrast, the latter's defense can be performed at a lower price, so it is widely used in applications such as large personal computers or monitors. Conventionally, the anti-glare treatment of the above image display device is achieved by bonding an anti-glare film to the surface of the image display device. The anti-glare film has been formed by, for example, using the following square wire 1 to apply a resin Μ with a fine particle size to adjust the film thickness @ to the substrate sheet i' so that the fine particles are exposed on the surface of the coating film, thereby making the table (4) 321900 4 The method of 201042295, etc. However, the use of the antiglare film produced by dispersing the above fine resin solution causes the arrangement or shape of the surface unevenness to be affected by the dispersion of the fine particles in the resin solution.

Ο The desired surface unevenness is obtained, and when the haze of the anti-glare film is set to be low, there is a problem that a sufficient anti-glare effect cannot be obtained. Further, when the above-mentioned conventional anti-glare film is disposed on the surface of the image display device, there is a problem that the entire display surface is whitened by the scattered light, and the display color becomes cloudy, which is called "white turbidity". Furthermore, with the recent refinement of the image display device, there is a possibility that interference occurs between the pixels of the image display device and the surface unevenness of the anti-glare film, and as a result, a brightness distribution is generated, and the so-called (flickering) phenomenon of the display surface is not easily seen. In order to eliminate _ although there is an attempt to scatter light by dispersing a difference in refractive index between the heterolipid and the fine particles dispersed in the resin, when the anti-glare film is disposed on the surface of the image display device, the ruthenium particles On the other hand, the transparent resin layer does not contain fine particles, and the anti-glare property is exhibited only by the fine unevenness formed on the surface of the transparent resin layer. For example, Japanese Laid-Open Patent Publication No. 2002-1891-6 discloses an anti-glare film in which a cured layer of an ionizing radiation curable resin layer is laminated on a transparent resin film, and the cured layer has a three-dimensional 10 point. The average particle size and the average distance between adjacent convex portions on the three-dimensional thickness reference surface respectively satisfy two predetermined fine surface irregularities. The anti-glare film is embossed by embossing The ionizing radiation-curable resin is cured and sealed in a state in which the ionizing radiation-curable resin is coated with the transparent resin film. However, the anti-glare is disclosed in Japanese Laid-Open Patent Publication No. 2002-189106 321900 5 201042295. In addition, it is difficult to achieve a sufficient anti-glare effect, and it is possible to suppress the whitening, the high-precision, and the suppression of the flicker. In the Japanese Patent Laid-Open No. Hei 6-34961, JP-A-2004-45471, and JP-A-2004-45472 The technique of using a film in which fine concavities and convexities are formed on the surface as a light diffusion layer disposed on the back side of the liquid crystal display, rather than an anti-glare film disposed on the display surface of the display device, is disclosed. In the method of forming irregularities on the surface of the film, the embossing roller filled with ionizing radiation hardening having a shape in which the unevenness is reversed is disclosed as disclosed in Japanese Patent Application Publication No. 2004-45472. The resin liquid is then brought into contact with the filled resin in synchronization with the rotation direction of the roll intaglio, and is placed when the transparent substrate contacts the roll intaglio A method in which the resin is hardened between the gravure and the transparent substrate and the cured resin and the transparent substrate are adhered to each other, and then the laminated body of the cured resin and the transparent substrate is peeled off from the light intaglio plate. In the methods disclosed in Japanese Laid-Open Patent Publication No. 2004-45471 and JP-A-2004-45472, the composition of the ionizing radiation-curable resin liquid that can be used is limited, and it cannot be expected to be leveled until diluted with a solvent. Therefore, it is conceivable that there is a problem in the uniformity of the film thickness. Further, in this method, since it is necessary to directly fill the resin liquid into the embossing roll, it is necessary to ensure the uniformity of the uneven surface. It is necessary to have a southern mechanical precision for the embossed roll intaglio, and it is difficult to make an embossing roll. Next, in the case of the method of making the film 321900 6 201042295 which is used for the production of the film having the unevenness on the surface, the use of the method is disclosed in the above-mentioned Japanese Patent Publication No. Hei 6-34961, the blasting, etc. Body, and then by its surface by electronic engraving, eclipsing the nickname of the bullet = the method of bumping. Furthermore, in Japan, the method of special opening, and the method of making a embossing roll embossing by the bead shot method is disclosed in Japanese Patent Laid-Open No. Hei. No. 04-90187. The step of mirror-grinding the metal woven layer, the step of embossing the surface of the metal (four) layer, and the method of making, smashing, and pulsing the roofing roll as needed. Next, there is a state in which the squeezing process is applied to the surface of the embossing roll as described above, and the distribution of the unevenness and the distribution of the unevenness, the anti-glare spray (4) the depth of the concave portion, and the reproducibility are excellently obtained* The problem of good bump shape. There is a technique for forming a concave-convex mold surface by a sandblasting method, which is described in the above-mentioned Japanese Patent Publication No. 2,02,189,106. Further, in the above-described mold surface in which the concavities and convexities are formed as described above, it is preferable to form a recessed surface and to improve the durability in use, and it is preferable to use a chrome-plated surface for the purpose of hardening and The content of corrosion prevention. In addition, in each of the above-mentioned Japanese Patent Publication No. 2004-45471 and Japanese Patent Application Laid-Open No. Hei No. Hei No. Hei. No. Hei. The technique of forming fine irregularities on the surface. In the embossing roll manufacturing method, since the hardness is high (four), the mouth is hit or the impact is made, it is difficult to form unevenness, and it is difficult to precisely control the shape of the unevenness. In addition, as disclosed in Japanese Unexamined Patent Publication No. Hei. No. Hei. No. Hei. No. Hei. No. Hei. The fine cracks generated by chrome plating have a problem that it is difficult to make a concave-convex design. Further, fine cracks are generated due to plating, and there is a problem that the scattering characteristics of the anti-glare film finally obtained do not change direction. Moreover, due to the combination of the material of the surface of the embossed rough base material and the combination of the seeding, various changes are made to the surface of the roll which is finally obtained. Therefore, in order to obtain the desired surface unevenness shape with high precision, it is also necessary to select an appropriate one. The pro-surface material and proper seeding issues. Further, even if a desired surface unevenness shape is obtained, depending on the plating type, there is a case where the durability is insufficient at the time of use. In Japanese Patent Publication No. 2000-284106, there is described a technique of performing a residual step and/or a film lamination step after performing a blasting process on a substrate, but a metal plating layer is provided before the blasting step. There are no records or hints. Further, JP-A No. 6-53371 discloses a technique of performing electroless nickel plating after polishing a substrate and performing sandblasting. Japanese Laid-Open Patent Publication No. 2007-18752 discloses a technique in which a base material is subjected to copper plating or mineral nickel, and then subjected to polishing and sandblasting, and then a bond is formed to form an embossed plate. And, in Japanese Unexamined-Japanese-Patent No. 7_237541, it is described that after the implementation of the ore or the town, the grinding is carried out and sandblasting is carried out, and then, after performing the remaining step or the copper ore step, the clock is implemented to make the health Pattern technology. In the production method using the above-described sand blasting process, since the surface uneven shape is formed in a difficult-to-tight control state, a large uneven shape having a surface concave-convex shape having a period of 50 or more is also produced. As a result, the concave-convex shape of the monochromatic image interferes with the pixels of the image display device, and the distribution of the meaning of 321900 8 201042295 degrees is generated, and the "flicker" problem in which the display surface is not easily seen is easily generated. SUMMARY OF THE INVENTION An object of the present invention is to provide an anti-glare treatment method for a transparent substrate, which can be used in an image display device to display a good anti-glare function and prevent identification due to white turbidity. The ability is reduced, and when used in high-definition image display devices, it can also exhibit high contrast without flicker. Another object of the present invention is to provide a method for manufacturing an anti-glare film which can exhibit good anti-glare function and prevent identification due to white turbidity when the film is disposed on the surface of the image display device. When the capability is reduced and it is placed on the surface of a high-definition image display device, it can also exhibit a high contrast without producing a flicker. Still another object of the present invention is to provide an image display device having an anti-glare property having the above display characteristics. Still another object of the present invention is to provide a method for producing a metal mold which is suitable for use in the above-described antiglare treatment method and method for producing an antiglare film. As a result of meticulous research, the inventors of the present invention have found out that the first pattern composed of image and image data is produced, and then used for the first pattern to at least remove or reduce a filter having a low spatial frequency component whose spatial frequency is not up to a specific value to produce a second pattern, and then processing the uneven shape on the transparent substrate according to the second pattern; by this method, good reproducibility can be processed The uneven shape is formed on the transparent substrate, and a sufficient anti-glare effect is exhibited, and white turbidity and flicker and a decrease in contrast can be sufficiently suppressed. Furthermore, with the above filter, 9 321900 201042295 it is found that a high-pass chopper and a band-pass chopper can be suitably used; the high-pass chopping line is removed or reduced by the spatial frequency component contained in the first pattern. ^ The lower limit value B' is a low spatial frequency component composed of a low spatial frequency, and the spatial frequency component is constructed by the lower limit value B and the above space _ rate (hereinafter, also depends on the lower limit value B) , the case where the W-resolved limit value β is referred to); and the pass-through device is removed or reduced by the spatial frequency lower than the specific lower limit value β in the air-rate component contained in the third 丨_ a low-resolution component and a high-altitude-off component composed of a spatial frequency exceeding a specific upper limit τ, and extracting a spatial frequency component composed of a spatial frequency of a specific range from the lower limit value β to the upper limit value ( ( Hereinafter, the lower limit value Β and the upper limit value τ of the specific range are also referred to as a spatial frequency range lower limit value Β and a spatial frequency range upper limit value τ. The present invention has been completed based on the above knowledge and insights and the review of the money. The anti-glare processing method for a transparent substrate provided by the present invention has a spatial frequency of a filter applied from a first pattern to a first pattern in which a plurality of dots are arranged in a random manner or in which a luminance distribution is arranged. The step of removing or reducing at least a low spatial frequency component having a spatial frequency that does not reach a specific value to form a second pattern; and the step of squeezing the concave and convex shape on the transparent substrate according to the second pattern. In the above filter, a down-pass filter that removes or reduces only a low spatial frequency component whose spatial frequency does not reach a specific value from the spatial frequency component included in the first pattern can be preferably used. Preferably, the high-pass filter removes or reduces a low-pass filter having a low spatial frequency component whose spatial frequency is less than 〇 1 # nf] from the spatial frequency component of the first pattern 3 . 321900 10 201042295 * Furthermore, in the case of the above filter, it is preferable to use the two-time ratio components contained in the second pattern to remove or reduce the low-space step rate of the spatial frequency that does not reach a certain value, and remove or A band-passing wave device that reduces a spatial frequency component whose spatial frequency exceeds a specific value and thereby extracts a spatial frequency component of a specific range. In the anti-glare processing method of the present invention, the band-pass filter is used to extract The spatial frequency lower limit value 0 1 in the above-mentioned specific range spatial frequency component is red or U1/zm-1 or more, and the upper limit value τ is preferably, where "m) is used to process irregularities on a transparent substrate. The resolution of the processing device used in the shape. Further, the upper limit value and the lower limit value β of the spatial frequency are preferably such that the following formula (1) is satisfied: - ° · 20 &lt; 2x (TB) / (T + B) &lt; 〇 &lt; go (1 In the first pattern described above, for example, a pattern formed by irregularly arranging a plurality of dots can be preferably used. Preferably, the anti-glare processing method of the present invention comprises the step of applying a dithering method to the second pattern to produce a third pattern converted into discretized information. In this case, the step of processing the uneven shape on the transparent substrate is carried out in accordance with the third pattern. In the case of the dithering method, the error diffusion method can be preferably used. Furthermore, the preferred embodiment in which the f 3 pattern is converted into the second-stage discretized information H in the (four) processing method of the present invention is that the conversion error is spread over 3 pixels by using An error diffusion method in a range of 6 pixels or less to create a third pattern. The anti-glare processing method of the present invention is preferably provided with: for the third pattern converted into 321900 201042295 two-stage discretization information, the Monte Carlo method is used to move the isolated black or white pixels to create the fourth pattern. step. In this case, the step of processing the uneven shape on the transparent substrate is carried out in accordance with the fourth pattern. Preferably, the step of processing the uneven shape on the transparent substrate comprises: forming a mold having a concave-convex surface according to the second pattern, the third pattern, or the fourth pattern, and transferring the uneven surface of the mold to the transparent substrate The steps above. The step of processing the uneven shape on the transparent substrate is preferably carried out using a processing apparatus that performs processing based on the discretized information of the third pattern or the fourth pattern. Further, the method for producing an anti-glare film according to the present invention includes a step of arranging a plurality of dots irregularly or arranging a luminance distribution! a pattern, the application filter removes or reduces at least a low spatial frequency component having a spatial frequency less than a specific value from a spatial frequency component included in the first pattern to form a second pattern; and processing the unevenness on the transparent substrate according to the second pattern Shape: Steps. In the case of the above filter, it is preferable to use a conventional chopper that removes or reduces a low spatial frequency component whose spatial frequency does not reach a specific value from the spatial frequency component contained in the second pattern. Preferably, the pass chopper is a high pass filter that removes or reduces a low spatial frequency component having a spatial frequency less than 从 from the spatial frequency components included in the first pattern. Furthermore, in the case of the above filter, it is preferable to use a spatial frequency component contained in the second pattern to remove or reduce a low spatial frequency component whose spatial frequency does not reach a specific value and to (4) reduce a high space in which the spatial frequency exceeds a specific value. Frequency 321900 12 201042295 Rate component, and thereby extract a bandpass filter of a specific range of spatial frequency components. In the method for producing an anti-glare film of the present invention, the spatial frequency lower limit value in the spatial frequency component of the specific range extracted by using the band pass filter is preferably 〇· 01 vnf1 or more, and the upper limit value. Preferably, it is below. Here, the D system has the same meaning as the above. Furthermore, the upper limit Τ and the lower limit 空间 of the spatial frequency are preferably such that the following formula (1) is satisfied: Ο ❹ 〇 .20 &lt; 2 χ (Τ-Β) / (Τ + Β) &lt;〇.8 〇(1) In the above-described first pattern, for example, a pattern formed by irregularly arranging a plurality of dots can be preferably used. Preferably, the method for producing an anti-glare film of the present invention comprises the step of applying a dithering method to the second pattern to produce a third pattern converted into the discretized information. In this case, the step of processing the uneven shape on the transparent substrate is carried out in accordance with the third pattern. By the dithering method, t Τ is more than ± (4), and the ' g 3 pattern is preferably a pattern of information converted into two-stage discretization. A preferred embodiment of the anti-glare processing method of the present invention is to produce a third pattern by using an error diffusion method in which the conversion error is spread by 3 pixels or more and 6 pixels or less. Preferably, the method for producing an anti-glare film of the present invention further comprises the step of moving the isolated black or white pixel to produce the fourth pattern by the third pattern of the information converted into the two-stage discretization. . In this case, the step 1S1 丄 ^ pattern is formed by processing the uneven shape on the above transparent substrate. 321900 13 201042295 The embossed shape includes: according to the second step, the step on the transparent substrate is preferably a mold surface, and the second, the third pattern or the fourth pattern is used to make the concave and convex shape The step of transferring the surface onto the transparent substrate is carried out by using a processing device based on a step on the transparent substrate, preferably using real information to perform a discretizing process of the fL pattern or the fourth pattern. .趸^ The present invention provides a method suitable for use in the above-described method of the present invention and a method for producing a method of glazing. The method includes: a substrate for a mold: a surface-implemented = mold-grinding step Forming a photosensitive resin on the surface to be polished: a photosensitive resin _ into (4); exposing the sensation pattern of the above second bribe and enamel to a photosensitive tree; and exposing the first s ^4 case, a photosensitive resin film of the third pattern or the fourth pattern is developed; a step of etching using a developed photosensitive resin film as a mask to form irregularities on the polished plated surface; a photosensitive resin film peeling step of the photosensitive resin film; and a second plating step of performing chrome plating on the lifted/outper surface. In the mold manufacturing method of the present invention, it is preferable that the unevenness of the uneven surface formed by the i-th etching step is etched between the photosensitive resin 瞑 peeling step and the second plating step to make the unevenness a gentle second "dream". The concave-convex surface to which the wrought chromium is applied in the second plating step is preferably a concave-convex surface of the mold transferred onto the transparent substrate. That is, preferably in the first $ ' 321900 14 201042295 The step of the surface of the concave and convex surface, and the direct application of the plating is applied to the concave and convex surface of the mold transferred to the transparent substrate. For the secret (four) shape (four) layer with ==== The invention provides an image display device which is provided with an anti-glare treatment on the surface of a transparent substrate provided by the image display device by the above-described anti-glare treatment method of the present invention; The anti-glare film of the present invention; the anti-glare film of the farmer. (4) Obtained by the county u method according to the present invention, the anti-glare treatment dazzle of the transparent substrate can be provided. When the image display device is used, the anti-glare can be displayed. = sex month, to prevent the use of white turbidity The discriminating power is reduced, and even in the case of a high-definition image display device, 'there will be no flashing and can be compared to the south. In addition, according to the present invention, the above-mentioned good display can be brought to the forefront. Jade reproducible money is formed on the through-the-day material, and then the ray is used to obtain the anti-glare treatment method and the anti-glare treatment method of the present invention. The transparent base anti-glare according to the present invention The processing method and the method for producing the anti-glare film can provide a video display device having the same display characteristics. [Embodiment] <Method for manufacturing a transparent substrate _ at the financial riding y> The following is a comparison of the present invention. The preferred embodiment (4) is a detailed description of the anti-glare treatment method for the transparent substrate and the method for manufacturing the ruthenium-nuclear film: in order to form a micro-specific distribution with a specific spatial frequency distribution on the transparent substrate: 321900 15 201042295 After the first pattern composed of a pattern of a plurality of dots of irregularity is irregularly arranged, at least the spatial frequency component included in the first pattern is removed; high-pass filtering of a low spatial frequency component of a specific value === Wave A second pattern is formed; and a convex shape is added to the transparent substrate according to the obtained second pattern. Further, as will be described later, the obtained second pattern is converted into a The third pattern of the discretized information or the (4) 4 pattern obtained by processing the isolated material contained in the binarized third pattern towel according to the Monte Carlo method, the concave and convex shape is processed on the transparent substrate. Thus, in the present invention The fine concavo-convex shape is formed on the transparent substrate according to the second pattern, the third pattern, or the fourth pattern. The means for imparting anti-glare properties to the transparent substrate or the means for producing the anti-glare film disperse the particles The method in the transparent substrate has long been known, however, it is possible to impart a unique surface by using the method of the present invention which uses a high-pass filter or a band-pass filter or the like to remove or reduce the pattern of low spatial frequency components. The anti-glare treatment of the shape, which is capable of suppressing low spatial frequency components, is not possible with the above-described conventional methods. According to the anti-glare treatment method and the method for producing an anti-glare film of the transparent substrate of the present invention, it is possible to produce a concave-convex shape on a transparent substrate with good reproducibility and to obtain a sufficient anti-glare effect, and to sufficiently suppress white turbidity. An image display device with reduced generation of contrast and contrast. Further, when the band pass filter is used, since the high spatial frequency component which is difficult to process the unevenness can be suppressed, the unevenness of the unevenness in the processing of the surface of the transparent substrate can be further improved. 321900 16 201042295 Here, the "pattern" in the "first to fourth patterns" refers to an arrangement in which images, video data, discretized information are arranged in two dimensions, or arranged in an opening of a board. k The above image data may be image data (raster image) in the form of a raster, or image data (vector image) in the form of a squint. A raster image is a list of dots that represent an image as a colored dot (d〇t). The color information of each point is stored as a numerical value in the raster image. There are various types of formats for storing the raster image described above, but in particular, in a general format, for example, a bit map can be cited. In terms of bit maps, 24-bit color bit maps of red, green, and blue intensities are respectively widely used in 8-bit depth; 8-bit gray-scale bit maps in which luminance is represented by 256-order depth of 256 steps . In terms of the format in which the raster image is saved, in addition to the bit map, a PNG (Portable) that belongs to the image data such as a compression algorithm may be used.

Various formats such as Network Graphics), TIFF (Tagged Image File Format), JPEG, and GIF (Graphics Interchange Format). 〇 In the vector image, the value is used to store the starting point coordinates (position) of the line, and if it is a curve, it is the bending mode, the thickness, the color, and the color of the surface surrounded by the lines. Information such as a set of such numerical data, or a radius of a circle, a central coordinate, and a vertex coordinate of a polygon are also included in the vector image. ' In terms of the format of the saved vector image, especially in the general format, DXF (Drawing Interchange File) and SVG (Scalable Vector Graphics) can be cited. However, in the present invention, the vector image is not limited to the exemplified form as long as it belongs to the above definition. Furthermore, 17 321900 201042295

The vector image is not limited to the second element, but may also be I. In the vector image, there is a closed circle three-dimensional information. The above-mentioned "arrangement of the openings arranged in the board ψ~2 polygonal arrangement can be performed in each of the L. The pattern in the present invention is not limited to the above, and the summer is changed. Discretization 2 is provided as an image or image resource. The second element of the discretized information is stored in a floating point (for example, a 64-bit floating point), and s ' can be enumerated as 32 bits. Meta-integer, unsigned 16-bit integer) "Integer (for example, with the symbol (production of the first pattern). Various forms. The first pattern can be from the above-mentioned pattern, or it can have a shade or a numerical value. : Any use in the pattern may include, for example, more specific image data throughout the entire image range (image data with a plurality of white dots and a plurality of black dots arranged on a black background); It is equipped with a pattern of brightness/knife with a plurality of patterns, a large change pattern, a second element arrangement of discretized information, and the like, and a high pass or band pass filter for the first pattern. When m is equal to the manifold (for this point will follow In the case of Fourier transform by optical means, it may be a plate provided with an opening. Moreover, the toner is locally attached to the photographic film (backsheet) or the transparent substrate on which the pattern is formed. It can be used as the first pattern. The point arrangement, brightness distribution and the configuration of the opening of the board in the image data can be regular or irregular (rand〇m), but in the spatial frequency domain, it can be obtained from the wide The viewpoint of having a quadrangular shape processing pattern having an amplitude and a low regularity is preferable, and it is preferable to form an irregular pattern. A plurality of 18 321900 201042295 points are irregularly drawn over the entire image range to be created to create the first image.鸯 鸯 , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , This life is, for example, the x coordinate of the center of the point. The 丫 is marked by the majority of the points of WYxR[2xro]. 11 in are all natural numbers to produce a pseudo-like; as long as it is linear, the same as the v, the upper, the lower Lu,

Xorshift ^, Knuth's subtraction random number generator algorithm, race rotation (Mersenne Twister), etc. 1 can be f + 廒Ο Ο random number generation method c length, you can use any pseudo-like random number of hardware to limit It is intended to be a random number, but by means of hot noise and other products, the body is to create a pattern of the second place that is irregularly arranged. The distribution may be round, polygonal, or the like, or may be a plurality of points of the same shape, or a plurality of the same shape of the same type of H. Therefore, the size of j can be a straight line (four) of all the points being the same, or a point c, which can be irregularly arranged with one type of point, and a plurality of points are used to make the first one. The pattern may also be such that it has a plurality of points having a plurality of dot diameters. The average point diameter of the points of the pattern) (the mean value) (the point diameter of all the points in the pattern is limited by the penetration, but when the band-pass 11 is used, the target has a peak of the spot diameter. And it is set to be lower than the penetration frequency and the low spatial frequency domain does not have a peak value, and is preferably 16 to 32 when the ultra-horse. When the average point diameter is eight, 3 has more Low spatial frequency that affects flicker

# The second pattern produced by the image produced unevenness in shading. On the other hand, the average dot diameter of the point constituting the first water is too small 'the irregularity of the first pattern when the amplitude of the spatial frequency component of the band pass filter 321900 19 201042295 =:: is small. Sexuality, but it is not possible to obtain a better map. The average dot diameter is preferably used, and the supply is given to the belt (4). The money G.5ΧΠ/(2ΧΤ))*. Thereby, when the dot filling rate bit * gate frequency range is satisfied, the frequency of the force extracted by the band wave device is sufficiently contained, and the second pattern which is less likely to cause unevenness in the shading is easily produced. When point 21 = the through-the-view device, (4) is set to have a peak value of ... in the range of the penetrating county and is not preferable in the low-frequency frequency domain of the range below the penetration band, and thus constitutes an average point. The diameter system is usually 4 to 5 mm, preferably one or more:: or more, preferably 32 Am or less, more preferably 3 Å or less, and particularly preferably 12 Å. When the average spot diameter exceeds the center m, it contains a lot of low ages that affect the flashing, and * is easy to produce unevenness in the second pattern produced. By arranging a plurality of dots to make the i-th pattern The filling rate (with the area occupied by the full-area midpoint) is preferably 20 to _, 2 to ^ is better, 30 to 70% is better, and 3 to 6 is better, 4 to _, for example, 0 may be around 50%.) When the number of dots is extremely small and the dot filling rate in the second pattern is less than 2G%, the second pattern generated may be characterized by concentric circles. (4) The unevenness formed is not able to obtain a better irregularity pattern. Further, when the dot filling ratio exceeds m, the unevenness of the closed circular pattern tends to occur, which may be impaired. Regularity. The first pattern can be produced in the form of image data in vector form. It can also be 321900 20 201042295 r==r etc::::., occasions can be used, one;;=^^: The method of breaking the fine pattern is high solution = Fig. 1 shows the manufacturing method of the anti-zero film which can be used in the transparent substrate of the present invention. An enlarged view of a preferred example of the first pattern produced by arranging a plurality of dots. The i-th pattern shown in the i-th image is a grayscale image of an 8-bit gray scale, and the black circular region is a dot. In the present invention, the diameter of the dot is referred to as "dot diameter", and the average of the dot diameters of all points in the pattern is referred to as "average spot diameter". The i-th pattern shown in Fig. i has a flat and mean point diameter of 丨6 from m. Furthermore, the image resolution is 12800 dpi. That is, the size of 1 pixel is equivalent to 2# m. In the second pattern shown in Fig. 1, the size of the image is WX = 0.512 mm, WY = 〇 · 512 mm, and the filling rate of the dot is about 50%. Further, the pseudo-number of the coordinates of the center point of the decision point is generated by supplying the value 607 to the SFMT verl. 3. 3 of the SIMD oriented Fast Mersenne Twister program installed by the group of Hiroshima University of Japan. Further, in the first pattern, it is preferable to use a pattern in which a luminance distribution is arranged, and for example, a raster image in which shading is determined by a random number is used. The concentration of each pixel (pixel) of the raster image is determined by a random number or a pseudo-disorder generated by the computer, whereby a pattern having a small regularity can be obtained. The method of determining the pixel density is exemplified by the case of using the pseudo-random number of the real number of 21 321900 201042295 in the range of 0 to 1. Although the gray scale of the pixel can be any 'but easy to handle, the gray scale depth has 1 bit, 8 bit, 16 bit, 24 bit, etc.' and preferably 8 bit (256 gray scale: index 0) To 255). For example, in 8-bit grayscale, for piXCEL[x,y] with 8-bit depth, an image can be generated by substituting PIXCEL[x,y] == r[x + yXimagewidth]x255. The pixel coordinates in the 'x, y image; the image width of the ImageWi dth system. In this example, although an image with an average index of 127 to 128 can be produced, an additional average image can be produced by an additional offset (0ffset). Fig. 2 is a view showing an example of a first pattern formed by a raster image in which shading is determined by random numbers. Fig. 3 is a view showing a part of the first pattern shown in Fig. 2 enlarged. The raster image shown in Fig. 2 is an 8-bit grayscale image which is determined by a pseudo-number to determine the brightness of 1 pixel and 1 pixel. Specifically, for a 2-dimensional array having an 8-bit depth PIXCEL[x , y], by substituting PIXCEL[X,y] = R[x + yx

ImageWidth]x255 is made. Among them, the pixel coordinates in the χ and y images; the image width is the pixel width of the X coordinate. In the case of permutation R[], the pseudo-like number obtained from the subtraction random number generator algorithm of Knuth is used, and the pseudo-random row is taken by Microsoft c〇rp〇rati〇n. The RandomDom NextDouble method included in Net FrameW〇rk2.0 Class Ubrary generates a value between 〇〇 and 1〇. Further, the first pattern may be a second element arrangement of discretized information generated in the same manner as the above raster image. In this case, in order to determine the value of each element of the arrangement, it is intended to be used. ^ 321900 22 201042295 ,, the wave of crying: the state can be based on, for example, the application of high-pass filter or bandpass == the device used to process the concave-convex shape on the transparent substrate Fan® and Gudan R Tian are in the wide-grating ~Η故&quot;&quot; It is better to use the random number to decide the bleak cry: "rr::iffiage". This is due to high-pass filtering = pass-through The range of the spatial frequency extracted by the wave device makes it easy to maintain the irregularity of the first pattern. Ο Ο The fourth picture is a comparison of the majority (4) produced by the irregular case (not a simple point _ The obtained second element (4) obtained by converting the spatial frequency distribution obtained by the high-speed Fourier, FFT) into the spatial frequency band and the first pattern of the raster image (random raster image) by the number of ILs The quadratic element is an example of a spatial frequency distribution obtained by converting the Fourier transform into a spatial frequency band, and displays the amplitude intensity in a region of a spatial frequency of 0.30/zm-1. As shown in Fig. 4, the irregular dot pattern is Compared with the random number raster image, the system is The spatial frequency 〇•l〇#m has a local amplitude intensity in the region. Further, it will be described later in detail. (Production of the second pattern) Anti-glare treatment method and prevention of the transparent substrate of the present invention In the method for producing a glare film, the second pattern is applied to a filter having a low spatial frequency component having a spatial frequency that does not reach a specific value by applying at least a spatial frequency component included in the ith pattern to the first pattern. In the present invention, it is preferable to use a high-pass filter as the filter, the high-pass filter only removes or reduces the spatial frequency from the spatial frequency component contained in the first pattern to a low value of 321900 23 201042295. The high-pass waver of the spatial frequency component is compiled into the data waver by removing or reducing a low spatial frequency component whose spatial frequency does not reach a specific value by the spatial frequency component of the first = S, and The spatial frequency component of the range of the high frequency component '(10)(4) whose spatial frequency exceeds a certain value is removed or reduced. A = σ 'pattern contains the spatial frequency component corresponding to the change.遽 or pattern contains components with high spatial frequency, and the variation is relatively small; or there are fewer components with higher spatial frequency in the more sparse pattern. By transporting (10) through the wave band, the ith pattern can be obtained from the ith pattern. The spatial frequency is f to remove or reduce the spatial frequency component of a specific range, that is, the low spatial frequency component of the flash period component. By using a high-pass filter or a bandpass chopper, the transparency can be reduced. a low spatial frequency component in the first, second, third, or fourth pattern of the thief on the substrate. By using a high-pass chopper or band-pass filter with respect to the second pattern' It can be implemented by the following = operation. Weekly string (1) is converted to spatial frequency domain - first, in order to be available from the first! The spatial frequency component of the pattern is extracted as a component of the floating point type (that is, when the specific low spatial frequency is removed or reduced to become a junction image, the first pattern is converted into the value of the prime value of the prime according to the need. The sub-arrangement /, the middle X y system displays the position on the vertical coordinate in the raster image, and the second element row obtained in the above manner is used by means of obtaining the various spatial frequencies in the first! ] 321900 24 201042295 == The spatial frequency component and the spatial frequency = the frequency distribution between the patterns. The size of the spatial frequency component is obtained &amp; and Lu's optical technique, mathematical technique, etc., generally speaking, mathematics The method of obtaining the method is widely used. The mathematical method of one m: the size of the frequency component is called Fu-Feng. The Fourier transform system can be used to make the discrete Fourier transform of the machine by using the knives. The following is referred to as (4). Therefore, the conversion to the spatial frequency domain can be performed by, for example, using the DFT of the second element in the arrangement of the elements obtained by the ith pattern by using a differentiator. - The DFTi寅 algorithm (algori) In the case of thm), although a generally known algorithm can be used, it is preferable to use the c〇〇ley_Tukey type algorithm because of its good calculation speed. The DFT by the Cooley-Tukey type algorithm is also called &gt; High-speed Fourier transform (hereinafter referred to as FFT). When the first pattern is produced in raster form, the image data of the raster form can be easily converted to a spatial frequency domain on the computer by using the above DFT algorithm. When the first pattern is created and converted into the spatial frequency domain by using the DFT algorithm, the image data in the vector form is converted into a raster form, and the image data converted into the raster form is converted into a primary element g[X, on the computer. y ], where X and y are the positions on the vertical coordinates in the raster image. When the first pattern is made into a general gray-scale image with an '8-bit gray scale, for example, 255 is assigned to the white area. The 〇 is assigned to the black area. By means of the DFT method and using the values, the image data is converted on the computer into a spatial frequency domain quadratic arrangement G[ fx, fy ], where 'fx, fy are separately displayed The number of spatial frequencies in the X direction, the number of spatial frequencies to the y square 321900 25 201042295. In addition, when the first pattern is arranged as a second element of the discretized information, of course, the DEF can be applied to the second element arrangement. The second element arrangement G[fx, fy] ° converted to the spatial frequency domain on the computer can also be used to perform the first pattern of the second element arrangement belonging to the discretized information, or to be converted into a quadratic element. The processing of subtracting the total element average value PA of the second element array from each of the array elements of the first pattern, for example, a first pattern of a gray scale image which is formed, for example, by an 8-bit gray scale having a value of 0 to 255. After conversion to the quadratic element arrangement, the process of subtracting the total element average value PA of the second element arrangement from each of the arrangement elements can be performed. When a gray scale image having an 8-bit gray scale having a value of 0 to 255 is converted into a quadratic array, a spatial frequency vector having an amplitude at a spatial frequency 获得 is obtained. This is due to the fact that all the elements constituting the second element arrangement are biased. In the antiglare treatment method and the method for producing an antiglare film which are applied to a transparent substrate, it is important to grasp the characteristics of the surface unevenness shape imparted by the transparent substrate, and the amplitude of the spatial frequency 0 is understood to be the final concave and convex shape. The characteristics are not meaningful information. In order to make the amplitude zero at the spatial frequency 0, by performing the process of subtracting the total element average value PA of the second element arrangement from each of the array elements, the characteristics of the finally formed uneven shape can be easily grasped. Fig. 5 is a view showing a binary spatial frequency distribution map obtained by converting a binary arrangement obtained by the first pattern shown in Fig. 1 into a spatial frequency band by FFT. In Fig. 5, the horizontal axis and the vertical axis show the spatial frequency. The point at which the two axes intersect is the point at which the spatial frequency is 0. As it moves away from the intersection (0 point), the spatial frequency becomes larger. In addition, the intensity of each spatial frequency is displayed in the concentration of the color. The intensity of the vibration is increased. The system ===image_converted to the spatial band is the second element information of Fig. 5. However, since: dimension .==:', therefore, when displaying the spatial frequency distribution, the value of ς::: is the average of the amplitude intensity of the transverse ratio: the inter-frequency distribution. The spatial frequency of the -dimensional dimension ο: The spatial frequency distribution of the second element shown in the tth fifth figure is the upper one! The T-th line graph shown in the figure is an inter-frequency frequency distribution diagram of the A-transmission (obtained by the FFT decomposition into the null_rate). In Fig. 4, the horizontal axis shows the flatness of the amplitude of the elements to which each spatial frequency of the spatial interpolation #H belongs. =W (4) (4) refers to the absolute value [x'fy]l of each element of the second element. Further, the average value is obtained by averaging the spatial frequency _ to the range of _ : = = = = = : : : : : : 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 The value of the spatial frequency to which the element belongs is determined by the value ia calculated by ...fy. The formula (Α) and the formula (Β) which are η belly and fa are expressed as follows. Fmax= (fxinax2 , fyD]ax2)i/2 (a) fa=(fx2+fy2)1/2 (8) refers to the maximum value of fx, and fymax refers to the maximum value of {. In the graph shown by the dotted line, even if the first pattern is created by a random irregularity, the first pattern has a peak shape in the specific space 321900 27 201042295. When there is a mixed peak (4), it may be impossible to obtain the desired two rooms because the south frequency of t is lower than the lower limit of the spatial frequency specified by || or the upper limit or lower limit of the band-passing wave The second pattern of frequency characteristics is therefore preferably corrected for the vibration of each element ' such that the amplitude of each spatial frequency is equal or substantially equal within a particular spatial frequency range. ... Figure 6 is not shown for the dashed line of Figure 4. An example of the result of the over-vibration correction in the spatial frequency distribution shown in the figure. The spatial frequency distribution before amplitude correction is shown by a broken line (same as the broken line in Fig. 4), and the spatial frequency distribution after amplitude correction is shown by a solid line. In the spatial frequency distribution shown in Fig. 6 / by correcting 'the amplitude of each element will be at the spatial frequency G to about 〇. 30#, s domain towel roughly - 疋. So, because of the high-pass filter or band The frequency band t of the button is set to a certain value, so the second pattern produced by the two-pass chopper or the band-pass chopper has a spatial frequency having a range of two Ingredients. The system is controlled by the wave H The pattern characteristic produced by the band-pass filter is more than itm H 'The above (4) correction is made by using the corrected; = the value c' is obtained by the formula: (4) 丨 '丨 real number ^ j vibration ^ g However, it cannot be depreciated. (2) The m correction system can only be performed in the range where 1 '1 is a non-G value. (2) Application of the same pass filter or band pass filter Next, the spatial frequency domain obtained by DFT Medium - Corresponding to the high-pass filter, wave H or band (four) wave machine operation | implementation of (10) the low spatial frequency components removed by the second _ 'for the 321900 28 201042295 high-pass chopper also known as high-frequency chopper, Low-Cut Filter, in the field of signal processing, has the effect of removing or reducing the components that do not reach the specified frequency. The operation corresponding to the high-pass filter refers to removing or reducing the spatial frequency of the first pattern. An operation of extracting a spatial frequency component composed of a spatial frequency equal to or greater than the lower limit value B' by a low spatial frequency component composed of a spatial frequency lower than the lower limit value B' of the spatial frequency range, and using DFT In the case of, more specifically, the phase In the conversion to the spatial frequency domain, the 0 is substituted (the amplitude is set to 0). The alignment element of the spatial frequency component (the real part of the conjugate amplitude) is lower than the range specified by the lower limit B' of the spatial frequency range. , each of the imaginary parts, or multiplied by an operation whose absolute value is much smaller than 1. The absolute value is much smaller than the value of 1 and is generally exemplified by the performance of the filter called a high-pass filter. The value is closer to 0, the absolute value is greater than 0. The value is closer to 0, the absolute value is closer to 0, or the absolute value is closer to 0. In general, the closer the absolute value of the multiplied value is to 0 (package (Q contains 〇), the more it becomes the ideal high-pass chopper. If the dependence on the spatial frequency of the penetration ratio of the high-pass filter corresponds to a certain spatial frequency as shown in Fig. 7, the value of the lower limit B' of the spatial frequency range can be regarded as the starting of the rise. point. On the other hand, when the penetration band rises slowly, the value of the lower limit value B' of the spatial frequency range is set to a spatial frequency showing the intensity of the peak intensity of the penetration band of 1/2. The spatial frequency range upper limit value T and the spatial frequency range lower limit value B of the band pass filter are also the same. The penetration ratio shown in Fig. 7 and Figs. 8 to 14 to be described later shows the absolute value of the value multiplied by each of the above elements. In addition, in '29 321900 201042295', the real number is multiplied to perform the operation of the corresponding band pass filter and the pass filter. a ^ , : the spatial frequency band (penetrating frequency band) extracted by the second skin device (4) The amplitude of the coded image (the ratio of the amplitude intensity before the X-pass filter) can be as shown in Figure 7 In the example shown in the figure 8 of the penetration band, the value is changed as in the example, and the frequency domain has a complex peak value. As shown in Fig. 9, it is called a band chopper. In the field of signal processing, the frequency of the microphone passes and the frequency other than this acts. The operation corresponding to the belt and the wave means the one obtained by the above method! In the frequency distribution between the patterns (4), the spatial frequency components contained in the (four) subtraction pattern are low spatial frequency components composed of frequencies below the spatial frequency and the high spatial frequency components composed of the spatial frequencies of the transmissions. * The door frequency is the meaning of the frequency, and the space frequency of the specific range of the upper limit is extracted: the operation of the frequency is "". In the case of using DFT (set _ to G) does not include (4) the empty coffee row that is passed: the limit of the element limit and the space frequency range lower limit (four) kg specified range of elements, or multiplied by the absolute value of the value far than i The small value is as described above. Small value operation. In the absolute, two: two by: pass the filter to extract the spatial frequency band (penetration band) intensity relative; ^ transport ratio (using the high-pass chopper after the amplitude, the same, the amplitude of the filter before the amplitude of the ratio The system can be as shown in Fig. 321900 30 201042295. The example of the figure (the shape of the peak of the penetration band is determined by the whole band, and can also be a branch). The value of the penetration value is Gaussian. Example of the case where the peak shape of the transmission band is also the peak of the transmission frequency domain of the space _ rate axis left outer (the penetration band peak - the shape is at the peak: it::= =

G, the tooth is also tribute to the two peaks, which can be composed of a plurality of peaks, and is also composed of a number of peaks, and also has a spatial frequency distribution as shown in Fig. 5. The quadratic spatial frequency after the ft graph _ The phase M1, the vertical axis, and the color density represent the same persuasion as in Fig. 5. For example, the corresponding band-pass filter is used to remove the spatial frequency component of a specific range specified by the __1_, or 』= intensity. Ο , 2 is the lower limit of the spatial frequency range given by the moon to the channel. The upper limit of the spatial frequency range T and the spatial frequency of the bandpass chopper: the ideal range of the limit B. The low spatial frequency component removed or reduced by the high pass filter or the band pass filter is preferably a low spatial frequency component corresponding to a spatial frequency below a period relative to the use of the antiglare treated transparent substrate obtained by the present invention. The image of the (anti-glare film, etc.) shows the 4-pixel size of the flat edge of the skirt (for example, when Na Na &lt; 3 colors are laterally juxtaposed, the average-edge pixel size of each of the threats is long and short The average value), a period of about 10 minutes or less. Thereby, the flicker of the image display device 321900 201042295 can be effectively suppressed. For example, the image display device specifically shows that when using a full-height image (resolution level, vertical 刖 point, etc. The maximum value of the spatial frequency of the low spatial frequency component of the pass filter or band pass filter, the lower limit of the frequency range of ::::, or the lower limit of the spatial frequency range: 佥曾=above. In addition, when used in an image display device equivalent to a diagonal of about 5 〇 to 3Γ2: 广χ平1366χ, vertical 768 points, etc.), the lower limit of the frequency range 6' or the lower limit of the spatial frequency range β Good 2 or above. From the same point of view, when the image display device corresponding to the diagonal angle is used, the spatial frequency range lower limit value ^ or the spatial frequency range lower limit value 8 is preferably 〇·〇3# ΠΓ1 or more. . When using a high-quality image display device equivalent to a diagonal angle of about 37, the spatial frequency range lower limit value B, or the spatial frequency range lower limit value B is preferably 〇〇4〆 when applied to the equivalent diagonal Approx. 2 (W high-definition image display, space frequency range lower limit B, or spatial frequency range lower limit Μ preferably 0.05# m1 or more. When applied to the equivalent of diagonally &amp; In the case of the enamel image display device, the lower limit value B' of the spatial frequency range or the lower limit value B of the spatial frequency range is preferably 〇〇7//mi or more. Thus, the display of the image by the corresponding image is displayed. The resolution and size, and appropriately adjust the lower limit of the spatial frequency range of the high-pass or band-pass chopper, so as to produce a low spatial frequency with or without the V relative to the scene display I. The pattern of the second, third or fourth component of the component is processed according to the pattern, whereby an anti-glare treatment for suppressing flicker can be achieved. 321900 32 201042295

Further, in the Shangtong filter, the spatial frequency range upper limit value τ is preferably 1/(Dx2) 〆 or less from the viewpoint of processing suitability. Among them, D ❹ Ο (/zm) is a resolution of a processing apparatus used for processing a concave-convex shape on a transparent substrate. When the upper limit value τ of the spatial frequency range is exceeded, it is difficult to provide a concave-convex shape on the transparent substrate because the processing reproducibility is difficult. The smaller the spatial frequency range upper limit value T is, the better the process reproducibility is, and the space frequency range upper limit value T is preferably 1/(Dx4) # ^ or less, more preferably xe) / /!!! 1 or less. When the spatial frequency range upper limit value τ is equal to or lower than 伽6)#mi, it is possible to use a highly productive laser scanning device to form a concavo-convex shape on a transparent substrate with good process reproducibility. On the other hand, the larger the upper limit value τ of the empty duty ratio range, the more the second pattern having a finer structure with two cycles is formed, and thus the process reproducibility is easily made difficult. The processing apparatus used for processing the uneven shape on the transparent substrate can be a conventionally known device, for example, a laser scanning device or a precision lathe can be used to expose the resist to a concave-convex shape using a laser scanning device. The point diameter corresponds to the resolution J) ( # m). In addition, when a precision lathe using a 4, sub-spherical ball mill is formed into a concave-convex shape, a ball mill having an end radius of r (am) is formed into a concave-convex shape, and the flat surface of each convex front surface and each position are added. When the angle formed by the face is within the radius (10 degrees behind the concave), 2xrKsin(0+18Ox;r)) corresponds to the solution being, for example, πι). Further, according to the second pattern, the mold weight D having the uneven surface is formed (the concave-convex surface is transferred onto the transparent substrate, whereby the concave and convex surfaces are processed and the mold is used to process the concave-convex shape on the transparent substrate. When the processing device is 7-shaped, the processing device used for the mold having the uneven surface. 糸' is produced 33 ^^19〇〇201042295 Furthermore, in the high-pass waver, in order to impart a suitable fine concave surface to the transparent substrate The length 'long period length 1/B of the shape 'belonging to the reciprocal of the lower limit value B of the empty_rate range B is the intermediate period length in the middle of the shortest period length 1/Τ of the reciprocal of the null_rate_upper limit value τ (1:Β +1/Τ)/2 is preferably in the range of 6"m or more 33&quot;. Ma 1 ηPer 1 〇d is equivalent to the _ length corresponding to the upper limit of the spatial frequency range given by the high-pass chopper (10) ) &quot; and the average value of the period length (10)) of the corresponding empty _ rate range lower limit B. When the pool (10) exceeds 33/zm, it is difficult to form a spatial frequency in the processing of the concave-convex shape on the transparent substrate. Lower than (U0W's fine concave and convex surface shape, and the beneficial method effectively shows anti-glare. When MainPe When the riQd is less than one time, in the processing of the uneven shape on the transparent substrate, it is possible to form a spatial frequency, and the fine uneven surface shape of 0.011, as a result, may be applied to a fine image display device (for example) When the obtained anti-glare film is disposed on the surface of the high-definition image display device, flicker is generated. As described above, the main purpose of using a filter such as a high-pass filter or a band-pass filter is to finally process the concave-convex shape. In the pattern (for example, the second, third, or fourth pattern to be described later), the low spatial frequency component composed of the spatial frequency lower than the lower limit B of the spatial frequency range or B is removed or reduced. (3) Production of the second pattern Next, the information of the spatial frequency by which the operation corresponding to the high-pass filter or the band-pass filter is performed is converted into a quadratic array by inverse discrete Fourier transform (IDft), and the second pattern is generated according to the second element arrangement. As for the IDFT algorithm, as in the case of the aforementioned DFT, the commonly known 321900 34 201042295 algorithm can be used. The second pattern can have 8-bit, 16-bit, 32-bit, Various bit depths such as 64-bit elements. Fig. 16 is an enlarged view showing an example of a second pattern created by applying a band pass filter to the first pattern shown in Fig. 1. Fig. 16 and Fig. 1 Similarly, the image data of 12800 dpi. The lower limit value B of the spatial frequency range and the upper limit value T of the spatial frequency range given by the band pass filter are 0.043^1^, 〇.059/zm-i, respectively, and 2χ( τ_Β)/(τ+Β) is 〇.3〇. When the second pattern is generated, conversion substitution may be performed so that the maximum value and the minimum value of the second element arrangement obtained by IDFT are respectively corresponding to the generation 2 The maximum/minimum value specified by the bit depth of the pattern. In other words, when the minimum value of the quadratic array element calculated by the resistance is set to j, and the minimum value is set to Imin, when the value of the element is converted into a pattern of 8 bits (〇_255), each of the patterns is substituted. The value of the pixel is calculated as + (Imax - Imm). The image data of the above-mentioned Fig. 16 is obtained by the above conversion. 〇 The above describes an example of a method of producing a second pattern by using a high-pass filter or a band pass filter of DFT, but the second pattern may be produced by the above method. For example, it is also possible to use a plate with a π-opening as a ^^1 pattern to perform Fourier transform on the plate by optical means, thereby obtaining 4 cases. Specifically, it is prepared to make the two-lens optical lens system through the two lenses, and the first ® case is placed in the first focus (four). At this time, the spatial frequency distribution of the image can be obtained from the focus of the two lenses (in the 'well' plane. In the Fourier plane, the penetration rate of the shoulder is spatially changed, so that the desired range of empty space can be obtained. 321900 35 201042295 Frequency penetration. The filtered output image is obtained from the focal plane on the opposite side of the Fourier plane of the second lens. For example, when the plate is configured such that only the center of the opening penetrates the Fourier face, only Obtaining a low spatial frequency component of the image as an output image. Conversely, when the center of the opening is shielded from light, only a high spatial frequency component is obtained as an output image. Therefore, by shielding the central portion and its peripheral portion from the Fourier surface, At the focal plane of the second lens, a second pattern having a desired spatial frequency distribution is obtained. (Conversion to Discretized Information and Creation of Third Pattern) The present invention is preferably the second obtained in the above manner. The pattern is converted into a pattern of discretized information. By making a pattern that is converted into discretized information, it is possible to create an additive suitable for processing the concave and convex shape. The pattern of the apparatus. For example, when the step of processing the uneven shape on a transparent substrate to be described later includes using a resist engineering such as a laser scanning device or NC machining (numerical control machining), the steps are used. The pattern is preferably a process of multi-valued by binarization, etc. In particular, when the step of processing the uneven shape on a transparent substrate to be described later includes using a resist process such as a laser scanning device, the second pattern is preferably Converted to information that is discretized in two stages, that is, preferably converted to a binarized pattern. This is due to the fact that the resist pattern is generated depending on whether or not the binary value of the laser is irradiated. By binarizing, images that can be used in laser scanning devices, etc. can be generated. In general, "discrete information" is called digital data, and the information processed on the computer is almost discretized. Information. In the case of discretization 36 321900 201042295, 5, the image data that can be processed on a computer such as a bi f map data, and 128-bit, 64-bit, 32-bit yuan , 16-bit 兀 各种 各种 各种 、 、 、 、 、 、 、 、 、 、 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 Or, the discretized information conversion with more stages will be used to convert the digital signal to the digital money expressed in less bit depth. For example, the information of the information of the scatter is exemplified by a cosine function in which the discrete function is a continuous function, and an 8-bit integer in which the number of decimal points of the 32-bit material is converted into a number of stages. The obtained high-pass filter or band pass filter and the filter have high continuity, so when a multi-valued pattern is obtained, and the binarized pattern is not used, it is preferably applied or banded under specific conditions. The second pattern is multiplied by the pass filter or the second pattern is multi-valued by a special method. The plurality of materials in the present invention. The heart is suitable for (1) binarization by the threshold method. For the method of binarizing the second pattern obtained by the operation of the band pass filter, the threshold method is preferably used. The 指标 order indicator (brightness value) sets a specific method by which white (or black) is applied to the gray illusion, and the pixel (color) of the value === is binarized. - - Black (or white 322900 37 201042295) The binarization by the second value obtained by the bandpass filter n is the binarization of the spatial frequency range and the lower limit of the spatial frequency range given by the bandpass filter. The upper limit value τ of the spatial frequency range preferably satisfies the following formula: ♦·0. 20&lt;2χ(Τ-Β)/(Τ+β)&lt;〇.8〇...〇) More preferably, the following formula is satisfied (2): 〇· 30^2χ(Τ-β)/(Τ+Β)^〇. 70 &quot;·(2). In the above formula (1) and (2) order 2χ(Τ-Β)/ (Τ+Β) is a numerical value of a range index of the spatial frequency of the second pattern extracted by the band-pass filter, that is, 2χ(Τ-Β)/(Τ+Β) is larger, the second The wider the range of spatial frequencies of the pattern, the smaller the range of the spatial frequency of the second pattern is smaller as 2 χ(τ_Β)/(τ+β). Figure 17 shows 2χ(Τ-β)/( The value of Τ + β) and the threshold value method are used to calculate the relationship between the maximum value of the autocorrelation coefficient obtained by binarizing the second pattern of the band-passing H. The maximum value of the autocorrelation coefficient is the maximum of the autocorrelation coefficient. The value of the autocorrelation coefficient is based on Wiener-Khinchin theorem. After the second pattern is converted into a quadratic element arrangement in the spatial frequency domain by quadratic Fourier transform, the coefficients of each element are squared, and the coefficient is obtained by inverse Fourier transform. The maximum value of the autocorrelation coefficient is displayed. The value of the self-phase relationship strength of the parallel movement itself is an index. Therefore, the higher the maximum value of the autocorrelation coefficient, the more easily the similar concave-convex shape is continuous in the uneven shape processed on the transparent substrate, although the concave-convex shape The period length is short, and it is easy to feel a specific periodicity when visually observed. The maximum value of the autocorrelation coefficient shown in Fig. 17 is the maximum value of the autocorrelation coefficient in the range of the moving distance of 20/zm or more. 321900 38 201042295 _ As shown in Fig. 17, it can be seen that the maximum value of the autocorrelation coefficient is extremely increased when 2χ(Τ-Β)/(Τ+Β) is 0.20 or less, and 2χ(τ_Β)/ '(10)) is above UG. At the time of 'maintaining a relatively low value. Therefore, in order to form a concave-convex shape that does not feel a specific periodicity on the transparent substrate, the recording of 2Χ(Τ-Β)/(Τ+Β) is preferably greater than G 2 ( ), more preferably G.3 () or more. 〇 On the other hand, as a result of the research, it is known that the second pattern has a wider range of spatial frequencies, and the second pattern is binarized by a threshold method by adding a plurality of components having different cycle lengths. It is easy to generate isolated small dots. Figure 18 shows the value of 2x(tb)/(丁+b), which is obtained by binarizing the second pattern obtained by the bandpass filter by the threshold method. The isolated small dots of the pattern produce a graph of the number. In the figure, "generating number" means that the center space frequency is set to 0.05 in the image obtained by binarizing the second pattern by the threshold method; απΓΐ, the uneven shape is processed on the transparent substrate. When the resolution D of the processing device (laser scanning device or the like) used at the time is 2 Am, the length of the continuous exposure range becomes the number of generated small dots having a resolution of 2×D//m or less. The existence of such isolated dots with a small number of consecutive elements hinders sufficient process reproducibility. In addition, the intermediate space frequency refers to the reciprocal of the above MainPeriod. As shown in Fig. 18, it can be seen that the larger the value of 2x(TB)/(T+B) is in the range of 2x(TB)/(T+B)4 〇8〇, the isolated small point There is a tendency for the number to be increased rapidly. On the other hand, when '2x(TB)/(t+b)4 〇^ or less, the number of isolated small dots is maintained at a relatively low value. The value of 2x(T~B;)/(t+b) is preferably not less than 0.80, more preferably 0.70 or less, because the processing reproducibility of the uneven shape is good. 321900 39 201042295 It is known that the spatial frequency range lower limit value B and the ★ gate frequency range upper limit value T preferably satisfy the above formula in order to form a concave-convex shape in which the process reproducibility is good and the specific periodicity is not perceived. (1), more preferably, the above formula (2) is satisfied. By using the band pass filter that satisfies the above formula (1) and more preferably satisfies the above formula (2), it is not necessary to perform the reduction processing of the isolated small dots by the Monte Carlo method described later. By performing the value-by-value method, the pattern with good process reproducibility can be obtained. Further, the spatial frequency distribution from the spatial frequency range lower limit value B to the spatial frequency range upper limit value T after the band pass filter is applied may be applied in the same manner as the spatial frequency distribution of the j-th pattern. Increase or decrease the processing so that the amplitude intensity becomes constant. By smoothly varying the amplitude intensity of the spatial frequency, a smoother concave-convex shape can be obtained. Here, in the resist engineering, when the ratio of the exposed regions is in the range of 30% to 7%, the suitability for money etching or development is good. It is preferably in the range of side to _. In order to make the pattern in which the second pattern is binarized by the threshold method satisfy the above condition, the threshold value must be set. This can be achieved by solving the (four) rate distribution for each pixel of the pattern of the younger brother 2, and using the value of the cumulative degree as the target ratio as a threshold value and binarizing. Specifically, it is as follows, for example. Figure 19 shows the slanting flute·| c. The image data shown in Fig. 16 is obtained by analyzing the frequency map of the gray scale index; (ρ β ^ β + seven dogs 13⁄4 days; f shows the cumulative rate Distribution map. From the accumulation rate shown in Figure 19, the eclipse is pi 1 , and the number of pixels below the gray scale index of 125 is 40%. Figure 20: The old mill', the cumulative rate distribution As a result of the analysis, the gray scale index 125 is used as the threshold value, and the face erosion value is calculated and the image data shown in Fig. 16 is binarized by the threshold 321900 40 201042295 value method. In the second pattern shown in the binarization, the filling ratio in the black portion (corresponding to the exposure region) is 40% by setting the gray scale index 125 to the threshold value. The 21st image display will be The 20th figure shows the spatial frequency distribution obtained by converting the binary element obtained by the binarized second pattern into a spatial frequency band by fast Fourier transform (FFT). As shown in Fig. 21, the The binarized second pattern has a low spatial frequency component that reduces the occurrence of flicker, and also reduces Since the spatial frequency distribution of the high spatial frequency component with reduced reproducibility is obtained, when the uneven shape is processed on the transparent substrate in accordance with the second pattern of the binarization shown in Fig. 20, excellent anti-glare performance can be expected. (2) Multi-valued method by binarization using a second pattern obtained by using a high-pass filter or a band-pass filter by multi-valued dithering Preferably, a dithering method is employed. In this case, the concave and convex shape is processed on the Q transparent substrate according to the third pattern obtained by the dithering method for the second pattern. The dithering method is used to convert the analog data into Digital data, or one of the methods of converting the bit rate or bit depth of digital data, can be positioned as one of the digital signal processing methods. Irregular signals are known by imparting a square probability density function or a triangular probability density function. 'There are various methods such as a method of reducing the deviation of the error when the signal is discretized, or a dithering method, an error diffusion method, etc. Among them, in the present invention, it is not easy to cause coloring due to ripple or interference. The repeated pattern of the cause, and the effect of suppressing the variation of the local average brightness can be expected, and further, by the optimization of the matrix, it is possible to suppress the occurrence of 41 321900 201042295, so it is preferable to make the _ diffusion] = the method of the dispersion method The feature is to make the enumeration generated in the discretization to convert the 8-bit 125 dragon ρ to the black-and-white bitmap image of the current gray scale to a 1-bit algorithm. For example, the Lai diffusion method The performance of the 'production pixel' (pixel) has a bright sigh. The pixel is converted to a 1-bit 2 gray-scale black and white bit map, 'must be converted to appear as an 8-bit The brightness value is 255 white, or appears as the black value of the brightness value 〇. The material system is converted to a closer value. Since the pixel with a 255 redundancy value of 64 is closer to 〇, it is converted to: a value of 0 (ie, black). At this time, since the conversion is compared with the image of the 8-bit gray scale, 'in the converted image, the luminance value error dispersion method means that the sum of the luminances of the pixels is reduced by 64. The miscalculation is based on the weight of the pre-determined decision to change the brightness value of the surrounding pixels to offset the resulting -64 luminance value error. The binarization is performed by repeating the above operations for all the pixels. As far as the weighting method is concerned, several matrices which are preferable in the field of image processing are known. For example, Floyd &amp;Steinberg; Jarvis, Judis and Nink; Stucki; Burks; Stevenson &amp;Arche; Sierra 3 Line; Sierra 2 Line; Sierra Filter Lite and other preferred weighted matrices. Fig. 22 is a diagram showing a diffusion weight map of the conversion error in the above exemplary matrix. As an example of a matrix, as illustrated by the example of Floyd &amp; Steinberg, pixel A is the pixel of the conversion object. As in the above example, 42 321900 201042295 because the conversion of pixel A (converted from brightness value 64 to 0), when the converted image will produce -64 brightness value error 'change by 7: 1 : 5 : 3 The brightness value of the adjacent 4 pixels is used to offset the brightness value error. That is, - the luminance values of the adjacent four pixels are increased by (7/16) χ 64, (1/16) x 64, (5/16) x 64, (3/16) x 64, respectively. Further, the pixel B of the hatched line of the hatched line indicates the pixel on which the binarization processing is completed. Further, the pixel described as "〇" is a pixel in which the weight of the error diffusion is not zero. Figs. 23 to 30 show an example of a third pattern obtained by applying the error diffusion method based on the matrix shown in Fig. 22 to the second pattern obtained by applying the band pass filter. The third pattern shown in Figs. 23 to 30 is created by the second pattern shown in Fig. 31 obtained by setting the gray scale image of the 8-bit gray scale, and is composed of a black and white image of 1 bit. The composition of the information. More specifically, the third pattern shown in FIGS. 23 to 30 is expressed at 12800 dpi with respect to a pseudo-random number line having a value of 〇 to 1 generated by the random number generator subtraction algorithm of Knuth. The first pattern of the octet image of the octet of the Q24匪4Q square, the lower limit value B of the spatial frequency range and the upper limit value T of the spatial frequency range are: B= l/MainPeriod*(l+ BandWidth/l〇〇) ..·(〇T=l/MainPeriod*(l-BandWidth/l〇〇)...(II) ' By using the error diffusion method of various matrices, the shape of the peak in the frequency domain is used. The second pattern shown in Fig. 31 obtained by the rectangular band pass filter is binarized, and is set to MainPeriod = 12 (in m) and BandWidth = 20 (%). In addition, in order to easily grasp the characteristics of the image, Figures 23 to 30 show the enlargement of a portion of the generated third pattern. 321900 43 201042295 Figure 32 compares binarization by various moment error diffusion methods shown in Figs. 23 to 3() A plot of the spatial frequency distribution of the third pattern after the third pattern and the spatial frequency distribution of the pattern after the centroid threshold method. If not shown in the figure, When the value method is binarized, the resulting pattern shows the amplitude intensity in the low space Γ = high. On the other hand, the error diffusion method is used to reduce the number. Therefore, the use of the error diffusion method can be used. = anti-glare treatment and anti-glare film for suppressing flickering. In addition, the pattern of == valued in Fig. 32 is the second pattern shown in Fig. 31, and the intermediate value of 1 word is used as the idle value. The value of the intermediate value m is set to white, and the value below the intermediate value of 127 is set to black, by the author. J clothes are so 'by the miscalculation of the error diffusion matrix known in the general drawing shown in FIG. Method; Uchida __ _ ' - Use, to obtain a third pattern with good spatial frequency characteristics. However, the method of making a third pattern that is binarized according to the error diffusion matrix will produce pixels of the same color and Most isolated pixels that do not exist in a certain number of groups (hereinafter referred to as "isolated points", the isolation = although similar to the above-mentioned "isolated small dots" conceptually, but as described later = meaning is not the same) The tendency to "isolated point" means existence A block (island) consisting of 16 or less consecutive pixels of the same color (pixels) in the 槿 赤. When the third pattern has many isolated points, there may be the following cases. There may be a block with one side of 4 pixels or less. (Island) requires extremely high precision for the embossing process based on the pattern such as the Up method or the wet process or the lathe processing, which hinders the reproducibility. 321900 201042295 The 33rd picture will be A graph comparing the number of generated isolated points generated by the third diffusion pattern by the error diffusion method known as the error diffusion matrix, and the comparison with the case where the threshold method is used. The numerical value shown in the figure shows the ratio of the number of isolated points generated when the pattern is binarized by the threshold method. As shown in Figure 33, even the Stonyson &amp; Arche matrix of the least generated point is produced 27 times more than the devaluation method, and when using Floyd & Steinberg's matrix, it also reaches 155: The inventors discovered after careful study that it is preferable to use a matrix that does not contain a short-distance error diffusion as an error diffusion matrix in order to suppress the number of generated points. Figures 34 to 42 show an example of an error diffusion matrix with diffusion distances of 2, 3, 4, 5, 6, 3+4, 4+5, and 3+4+5, respectively. These figures show the weights of the spread of the conversion error as in Fig. 22. The diffusion distance refers to the Sanskrit value error generated by the conversion of the white or zero of the pixel (pixel A) of the conversion target. The distance between the pixel that changes the luminance value and the pixel of the conversion pair Q image is 'diffusion distance 1'. The pixel whose brightness value is changed is adjacent to the pixel of the conversion target (refer to Fig. 34). "Diffusion distance 2" refers to a pixel whose pixel is changed from the pixel to be converted as a pixel whose luminance value is changed (j' pixels are interposed between the pixel whose luminance value is changed and the pixel of the conversion target) (refer to 35)). The diffusion distance of 3 or more is also the same. The "matrix of diffusion distance 3 + 4" in Fig. 40 refers to the combination of "matrix of diffusion distance 3" in Fig. 36 and "matrix of diffusion distance 4" in Fig. 37. The same applies to Figures 41 and 42. Figs. 43 to 51 respectively show an example of a third pattern obtained by the error diffusion method using the matrix shown in Figs. 34 to 45 3219 〇〇 201042295 42 . The second pattern used is the pattern shown in Fig. 31. Further, in order to easily grasp the characteristics of the image, Figs. 43 to 51 are shown enlarged by a part of the generated third pattern. Furthermore, in the 52nd figure, the number of isolated points generated when the third pattern is created by the error diffusion method according to the error diffusion matrix shown in FIGS. 34 to 42 is compared with the case where the threshold method is used. Picture. The numerical values shown show the ratio of the number of isolated points generated when a pattern that is binarized using the threshold method is produced. As shown in Fig. 52, when the error diffusion distance is 1, compared with the threshold method, an isolated point that reaches 247 times is generated, but as the error diffusion distance is set to be large, the number is generated. cut back. In particular, when the error spread distance exceeds 1, the number of isolated points will decrease sharply. As can be seen from the results shown in FIG. 52, in order to more effectively suppress the generation of isolated points, the error diffusion distance is preferably more than K, that is, the conversion error is spread over a range of more than 1 pixel, and the following is the same, more preferably 2 or more, more preferably 3 or more. Further, although the upper limit of the error diffusion distance is not particularly limited, it is, for example, 6 or less. Among them, a pattern produced by using a matrix having an error diffusion distance of 3 or more has a wide processing range, and good processing suitability can be expected. Fig. 53 is a binarization of the spatial frequency distribution of the third pattern of the 43th to 51st graphs binarized by the error diffusion method of the error diffusion matrix shown in Figs. 34 to 42 and the threshold value method. A comparison of the spatial frequency distribution of the pattern. The pattern that has been binarized by the threshold method is the same as that of Fig. 32. It can be seen from Fig. 53 that the vibration of the low spatial frequency component can be reduced compared to the threshold method, regardless of the use of any of the error diffusion matrices. (production of the fourth pattern)

The pattern that is converted into two-stage information (binarized) by the threshold method or the dither method has a situation in which many isolated points are included. At this time, the operation of reducing the isolated point can be repeated for the binarized pattern of the third pattern or the like to create the fourth round. At this time, a concave-convex shape was added to the transparent substrate in accordance with the obtained fourth pattern. By exercising the operation of reducing the isolated point, it is possible to process and then hide to form (four) impurities on the (four) material. The binarized pattern used for the production of the fourth pattern may also be a pattern that has been binarized by the threshold method, or may be binarized by a dithering method such as an error hybrid method. However, as described above, when the second pattern is created by using the band pass filter satisfying the above formula (1) and more preferably the above formula (2), the above-described isolated point reduction processing is not necessarily performed. In order to reduce the above-mentioned isolated point, it is preferable to use a method of moving a black or white pixel existing in a purely raised point of a binarized pattern such as a third pattern by Monte Carlo method. The method of the block of the same color (island). The Monte Carlo method is a general term for the simulation method based on random numbers. As far as the treatment of isolated points is concerned, the method of simply removing the isolated points is the most pure. However, when this simple method is used in image processing, there is a case where the average degree of exemption = local change, which causes an increase in the spatial frequency component. The squatting card is an effective way to deal with isolated points without locally affecting the average brightness. Hereinafter, a specific example of a method of processing an isolated point using the Monte Carlo method will be described with reference to the drawings. First, it is determined whether or not the target pixel (pixel) is an "isolated point". Here, the "isolated point" in the specific example of 321900 47 201042295 is different from the above definition, and is defined as the pixel of the same segment (the same color) as the target pixel among the nearest 8 pixels. The number is two or less. For example, when the target pixel is black, if the number of black pixels of the closest 8-pixel towel is two, it is determined to be an isolated point. The same is true for white pixels. Then, the pixel moves to the left of the nearest 8 pixels, and the heart is close to 5 (a). If the target pixel is black, it is determined to be an isolated point because the pixel is shifted by 1 pixel. _ pixels. In the 54th welcoming pixel, 2 of the 8 pixels of t is selected by random number: When the target pixel is black, the closest is moved to the empty most volume, black, so it is determined to be an isolated point, the object The pixel is connected to the pixel selected by the random number in the nearest 6 pixels in Fig. 54 (4). The 3 pixels in the prime are: in the right, when the prime is black, the closest image moves. ",, 邑, so it is not judged to be an isolated point, and the target pixel is not created by the reverse, and the operation of the Monte Carlo method is effectively reduced. The geo-carlo method repeatedly performs, for example, Μ to 60 frequency components = the following pattern can be obtained 'that is, when the gap between the vacancies of the band-pass filter is passed, the + gate frequency value is converted into a period of 3 pixels to 6 images. It is detected that the isolation is good, and the processing suitability can be expected. 55) The fourth pattern t) to (1) shows the graph caused by the number of applications of the Monte Carlo method. The pattern shown in Fig. 55 (a) to (f) For the 321900s, 2010, 2010, 2,295, 0, Π:, 〇 J pattern (diffusion distance 5), the Monte Carlo method is used to treat the isolated points by using 56 1 〇 and 6 分别 respectively. In addition, the third 卡罗 = Carlo method The relationship between the number of times and the number of isolated points is shown by the number of isolated points. The number of points is compared with the 33rd and 52nd values. The second pattern shown in Fig. 31 is used to create the pattern of the threshold method. The ratio of the number of isolated points. So, borrow

Ο = Repeat the Monte Carlo method to create a fourth pattern that reduces the number of isolated points and can expect better processing. In addition, the production of the fourth case is based on the use of the bandpass filter and the wave device in the case of the younger m, and the producer uses the second pattern. However, even if the second pattern is used by using the high-pass passer, In the case of the band pass chopper, the fourth pattern is obtained with good processing suitability by reducing the low spatial frequency components by the binarization and the reduction of the isolated points. The method for producing a pattern for processing a concave-convex shape on a transparent substrate as described above is applied to a second pattern by a dithering method (the towel is a mistaken method), and a Monte Carlo is produced. In the method of producing the fourth pattern by using the third pattern, when the second pattern is produced, the pattern of the low spatial frequency component and the isolated point can be obtained without using the band pass filter that satisfies the above formula (1). Therefore, the method is one of the preferred embodiments. (Processing of the concavo-convex shape according to the pattern) In this step, according to any of the patterns obtained as described above (the second pattern or the second pattern is converted into a two-stage information by the threshold method (two values) The patterned pattern, the third pattern, or the fourth pattern) is formed into a concave-convex shape on a transparent substrate, and imparts anti-glare properties to the transparent substrate. Concretely, 321900 49 201042295 The uneven shape is processed in accordance with the pattern by, for example, the following method. The processing device used for processing the uneven shape on the transparent substrate may be a conventionally known device, and for example, a laser scanning device, a laser processing device, a precision lathe, or the like can be used. As the laser processing apparatus, various processing apparatuses such as a laser marker, a laser engraving machine, and a laser processing machine can be used. The processing of the concavo-convex shape on the transparent substrate is preferably carried out by a processing apparatus that performs processing in accordance with the discretized information of the pattern. The processing apparatus for processing based on the discretized information is specifically a variety of NC processing apparatuses such as a precision lathe, an automatic engraving apparatus, a laser processing apparatus, and a laser scanning apparatus. In the case of a processing device, when using, for example, a laser scanning device, the discretization information is preferably discretized into two stages of information. When the concave-convex shape is processed by the above-described apparatus by discretizing into two-stage quadratic elements, it may be carried out as follows. First, the pattern is converted into a quadratic arrangement g[x, y] according to the brightness information. Here, X and y represent position coordinates indicated by each element of the second element array. Next, it is confirmed that the values of all the elements stored in the two-dimensional arrangement g[x, y] discretized into two stages are confirmed. Here, by discretizing into a two-stage operation, it is assumed that 0 or 1 is stored in the secondary element arrangement. In the processing of the concavo-convex shape, for example, when the value of the element g[al ' bl] stored in the second element corresponding to the specific position x = al, y = bl is 1, the laser is irradiated to the corresponding al in the processing apparatus. The coordinates of bl to form a recess. When the stored value is 〇, the corresponding coordinates are not irradiated to the laser. By repeating this operation for all the elements, the concave and convex shape can be obtained from the pattern. When the laser has an intensity capable of forming a concave portion in the object to be processed, a concave portion is formed by irradiation of the laser. When the laser intensity is weak, 50 321900 201042295 can also be used to sensitize the resist by laser scanning, and after the development, the concave portion is formed by rhyme, thereby processing the concave and convex shape. Further, when the pattern for the concavo-convex processing is composed of the second-element arrangement of the discretized information, the concavo-convex processing which is performed in accordance with the value of the two-dimensional system can be converted according to the characteristics of the processing device. The value, which is used to process β, for example, 'on a laser plus 1 machine or a laser engraving machine, can also be regarded as the number of laser shots. In the precision lathe _ control tool depth processing device, it can also be converted to; 7 pushes the amount of f. Hereinafter, the conversion of the values will be specifically described by taking the case of using the two-dimensional arrangement of the 8-bit 7L gray scales. At this time, it is assumed that the secondary element arrangement g[x, y] is a value of 0 to 255. The strength of the anti-glare property can be controlled according to the height difference of the uneven shape. The formula converted to the amount corresponding to the tool pushing amount is determined by the required height difference and the maximum and minimum values stored in the second το arrangement g[x' y]. When the height difference is set to 1, the tool pushing amount from the surface of the flat object of the coordinate χ and y is set to z, and Z = ^g ([x' y] - minimum Value) / (maximum_minimum) χ high and low difference meter 'to determine the tool push amount ζ. In the specific example described here, the tool pushing amount z is calculated by Z = g([x, y] - 〇) / (255 - 〇) xi^m. That is, when the value of g[x, y] is 255, the tool pushing amount Z is set to 1_, and when the value of g[x, y] is 0, the tool pushing amount 2 is set to $ 〇 _. The concave-convex shape is formed by repeating the steps of all the elements stored in the second element arrangement g[x, y]. When the depth is controlled according to the number of irradiations of the laser 51 321900 201042295, as long as the relationship between the number of times of irradiation and the depth of processing is confirmed in advance, the number of times of irradiation can be determined so as to correspond to the value of z. = The information of the pattern is converted into The transparent substrate is engraved into the depth of the -bein and reflected into the concave-convex shape, or formed into a concave according to the information of the pattern. P, or whether to form a concave portion, thereby processing the concave-convex shape. Further, when the height difference is excessively large due to the limitation of the resolution of the processing apparatus, the height difference can be reduced after the processing by etching the entire surface. Further, in the method of forming the concavo-convex shape of the transparent substrate, the method of directly performing the above-described processing on the transparent substrate may be employed, but it is preferable to adopt the following method: forming the concavo-convex shape according to the pattern in the mold by the aforementioned method Thereafter, the uneven shape of the mold is transferred onto the transparent substrate, and the uneven shape is formed on the transparent substrate. According to the pattern of the component 'βρ 4 main ϊι丨ΚΗ _ also 丨...

A ^ , Myt -pg 35^ Λγν -t. -w HH Ή- _L I A ~ π must be double-inch to listen to the surface of the substrate directly, and as far as the transparent substrate, as long as it is made of optically transparent material Bridge 俾, gp fer branch β丨丨ΪΓΗ 4*, 丨 成 成 成 成 成 成 成 成 成 成 成 成 成 成 成 成 成 成 成 成 成 成 成 成 成

On the surface of the device, an image display device that exhibits turbidity and flicker is obtained. The anti-glare film is disposed in the anti-glare property of the image and is highly effective: 321900 52 201042295 In the anti-glare treatment method and the anti-glare film manufacturing method of the transparent substrate of the present invention, the precision can be better and the processing can be reproduced. More preferably, the fine uneven surface shape is formed on the transparent substrate, and the productivity is also good. Therefore, it is preferably included. The following steps: according to any pattern of the above pattern (the second pattern or the second pattern is thresholded) Converting the pattern into a two-stage information (binarized), the third pattern, or the fourth pattern) to form a mold having a concave-convex surface (fine uneven surface shape), and transferring the uneven surface of the manufactured mold to Transparent base coffin. By peeling off the transparent substrate on which the uneven surface is transferred from the mold, a transparent substrate (including an anti-glare film) having a fine uneven surface shape can be obtained. The step of transferring the shape of the mold to the transparent substrate is preferably carried out by an embossing method. In the case of the ray method, the uv embossing method using a photocurable resin and the thermal embossing method using a thermoplastic resin are exemplified, and from the viewpoint of productivity, a UV embossing method is preferred. The UV embossing method forms a photocurable resin layer on the surface of a transparent substrate, and teaches the photocurable resin layer to be pressed against the uneven surface of the mold while hardening it, thereby transferring the uneven surface of the mold to A method of photocurable resin layer. Specifically, the ultraviolet curable resin is applied to the transparent substrate, and the ultraviolet curable resin to be applied is adhered to the uneven surface of the mold, and ultraviolet rays are irradiated from the transparent substrate side to cure the ultraviolet curable resin, and then The cured substrate on which the cured ultraviolet curable resin layer was formed was peeled off from the mold, whereby the shape of the mold was transferred to the ultraviolet curable resin. When the UV embossing method is used, the transparent substrate may be a film which is substantially optically transparent, and examples thereof include a cellulose triacetate film, a poly 53 321900 201042295 acid vinegar film, and a polymethyl group (four). Acid s|| Membrane, a normous ring-shaped compound having a norbornene-based compound as a monomer, a solvent for a thermoplastic resin, a film or a thin film (4): after the use of the υν embossing method The species of the ultraviolet curability tree is not particularly limited, and those suitable for commercial use can be used. Further, it is also possible to use a photoinitiator which is appropriately selected as an external curable resin, and to use a resin which is also hardened by visible light having a longer violet length than ultraviolet light, specifically: trimethylol alone a polyfunctional acrylic acid vinegar such as propane triacrylate acrylate, pentaerythritol or diacid vinegar, or a mixture of these compounds, and a suitable mixed use (10) E g〇7 (Ciba = τ (Swiss)), IRGAC brain 18 Price plus chemical company (Switzerland); &quot; Lucnnn TPQ (manufactured by BASF) photopolymerization initiator. The transparent substrate formed by the illusion is pressed against the mold. In the case of using a convex shape, as long as it is substantially transparent, == transparent substrate polymethyl methacrylate methyl vinegar, polycarbonate hydrazine, poly-pair such as cellulose can be used, and yin A solvent-cast film, an extruded film, or the like, which is a monomeric amorphous cyclic polyene, such as an ester or a triacetic acid hydrocarbon. The fat film can also be suitably used as a transparent substrate which is coated with the ultraviolet curable resin in the above-mentioned embossing method. δ (Manufacturing method of the board) Hereinafter, the method of the mold which can be applied to the manufacturing method of the present invention is described. The method of m ^ processing, the method of the anti-glare film ▲ $57 is a schematic display of the present invention 321900 54 The preferred part of the first half of the method of mold making in 201042295 is shown in Fig. 57. Fig. 57 is a schematic view showing the mold (4) of each step. The manufacturing method of the hair mold basically includes [1] the first ore cover. Step, (2) polishing step, (3) photosensitive dendritic film formation step, [4] exposure step, [5] development step, [6] first side step, [7] photosensitive resin film peeling step, [8] The second bonding step. The following describes the steps of the mold manufacturing method of the present invention in detail with reference to the 57th ® [i] first plating step. In the mold manufacturing method of the present invention, first, the substrate used in the mold is used. The surface is coated with a shaft or nickel plated. Thus, by the surface of the substrate for the mold

If the surface is applied to Lai or Lai, the adhesion or the light of the second coating of the second plating step can be made (4). "卩, as in the above-mentioned prior art, when the surface of the iron or the like is plated, or after the surface of the ore is formed by the sandblasting method or the beading method, the surface is made to have a secret time, and the surface is filled with (4) thick seams. It is difficult to make the surface of the mold concave and convex. For this, firstly, by applying Lai or Lin to the surface of the substrate, the above-mentioned monthly opening y can be excluded. The coating is high and smooth, and I: = the surface of the substrate, such as tiny irregularities or cavities, etc., because of the characteristics of the copper or nickel plating. The cerium plating which is described later in the center of the substrate and in the step of chrome plating can also eliminate the copper or yt coating of the chrome-plated surface due to coarseness, irregularities or cavities. Therefore, the reduction of fine cracks is the copper or the recording used in the first ore-covering step, except that it can be an alloy of 321900 55 201042295 pure metals, so it can be too much copper-based alloy or And nickel, the "steel" in the specification contains copper and copper alloys, respectively, using electrolysis = The meaning of containing brocade and recording alloy. The key copper or record can also be usually used in the heart:: into the playing, can also be carried out by electroless plating. When the bottom surface is nickel-plated, the 'money coating is too thin, and the upper limit of the thickness cannot be completely excluded. It is preferably 5 (um or more. Money coating..., limit, but in view of the fact that it is formed in the manufacturing method, it is suitable for the substrate for the mold, etc. by the treatment = two | by cost From the point of view, it can be exemplified that Ming, Tie Shao or iron can use light weight 18 respectively. Among them, the alloy of the body. ', ', ,, metal 'in addition to this 夕 卜 ' can also be used or The shape of the base material for the mold is mainly a two-plate type, or a cylindrical or cylindrical roll which has been conventionally used in the field. Anti-glare 2: The mold has the following advantages: The anti-glare film can be manufactured in a continuous roll shape. [2] Grinding step, clock 'In the grinding step, the substrate table is subjected to the rubbing in the first iron covering step After this step, the surface of the substrate = is ground to a state close to the mirror surface m (four) under the immortal: in order to, the essence of hope The degree is mostly applied to a metal plate or a metal roll as a substrate, such as cutting or grinding, so that the surface of the substrate remains processed 321900 56 201042295 traces even in the case of copper plating or nickel plating. There is a case where such processing marks remain, and in the plated state, the surface does not necessarily become completely smooth. That is, even if the surface described later is left with such a deep processing mark or the like, there is a processing mark. When the unevenness is deeper than the unevenness formed after the execution of each step, there is a possibility that the processing marks and the like remain, and when the anti-glare treatment is performed by the mold or the anti-glare film is produced, the optical characteristics are caused. Fig. 57(a) is a schematic view showing the surface of the flat-shaped mold substrate 7 subjected to copper plating or nickel plating in the first plating step (plating formed in this step). The copper or nickel-plated layer is not shown)' and is further mirror-polished by the grinding step, and has the surface 8 state. ❹ Regarding the method of grinding the surface of the substrate to which copper plating or nickel plating is applied, § ' There is no particular limitation. 'A mechanical polishing method, an electrolytic polishing method, or a chemical polishing method can be used. For the mechanical polishing method, a super finishing method, a polishing method (laPPing) method, a fluid polishing method, and a polishing method are exemplified ( Buffing) Grinding method, etc. Surface roughness after grinding according to JIS B 0601, "Line average roughness Ra is preferably 0. Below, more preferably 〇.〇5# The center line average roughness after grinding When the degree is larger than g. ^, the degree of rough seam: two after: T has the surface of the concave and convex shape, the surface after grinding is not particularly limited, by the processing time or the lower limit of force, ^ its limit, so there is no mosquito net From the viewpoint of t, the natural [] photosensitive resin ruthenium forming step is carried out, and in the photosensitive resin yttrium forming step, the substrate 7 for the mold after the mirror polishing is applied by the above-described polishing 321900 57 201042295 step is ground. The surface 8 that has passed through is coated with a solution in which a photosensitive resin is dissolved in a solvent, and heated/dried to form a photosensitive resin film. Fig. 57(b) is a view schematically showing a state in which the photosensitive resin film 9 is formed on the polished surface 8 of the substrate 7 for a mold. For the photosensitive resin, a conventional photosensitive tree bet can be used. In the case of a negative photosensitive resin having a hardening property, a monomer or a prepolymer or a diazide compound having a propylene acrylate or a mercapto fluorenyl sulfonate in a molecule may be used. Mixture of bisazide with diene rubber, polyvinyl cinnamate compound, and the like. Further, in the case of a positive photosensitive resin having a property in which a photosensitive portion is eluted by development and only a photosensitive portion is not left, a resin such as a phenoline resin or a phenolic resin can be used. Wait. Further, various additives such as a sensitizer, a development accelerator, an adhesion modifier, and a coatability improver may be blended as needed in the photosensitive resin. The photosensitive resin is applied to the ground coating 8 Bf solvent of the substrate 7 for a mold and applied. As the solvent, a 2-ethoxyethanol-based roller, a propylene glycol-based solvent, an ester-based solvent, an alcohol-based solvent, a ketone-based osmotic agent, a polar solvent, or the like can be used. For the method of coating the photosensitive resin solution, it is possible to use a meniscus coating, a fountain (f_tain) coating, a dipping coating, a spin coating, a light coating, a metal bar coating, an air knife coating, a knife. The method of strengthening Ning and the curtain. Preferably, the thickness of the coating film is after the drying is ^ 58 3219 〇 0 201042295 [4] exposure step, in the exposure step, the first pattern is made by using a high-pass filter or a band pass filter. The second pattern or the second pattern is converted to a pattern which is discretized into a two-stage (binarized) information by a threshold method. The third pattern or the fourth pattern is exposed to light in the photosensitive resin film forming step. On the resin film 9. The light source used in the exposure step can be appropriately selected depending on the photosensitive wavelength or sensitivity of the photosensitive resin to be applied, and the like, for example, a g line of a high pressure mercury lamp (wavelength: 436 nm) and a h line of a high pressure mercury lamp (wavelength: 405 nm) can be used. ), i-line of high-pressure mercury lamp (wavelength: 365nm), semiconductor laser (wavelength: 830nm, 532 nm, 488 nm, 405 nm, etc.), YAG laser (wavelength: 1〇64nm), KrF excimer laser (wavelength) :248nm),

ArF excimer laser (wavelength u93nm), F2 excimer laser (wavelength j57nm), and the like. In order to form the surface uneven shape with high precision in the method for producing a mold of the present invention, it is preferred to expose the pattern on the photosensitive resin film in a state of being precisely controlled in the exposure step. In the method for producing a mold of the present invention, in order to accurately expose the pattern on the photosensitive resin film, it is preferable to arrange the image according to a pattern formed on a computer or a binary element of discretized information. The pattern is scanned on the photosensitive resin film by using laser light from a laser controlled by a computer. In performing this 'laser scanning, a laser scanning device for printing plate production can be used. The laser scanning device may, for example, be a laserFX (manufactured by TMnk Laboratory). Fig. 57(c) is a view schematically showing a state in which the pattern is exposed to the photosensitive resin film 321900 59 201042295 9. When the photosensitive resin film is formed of a negative photosensitive resin, the exposed region 10 is subjected to a crosslinking reaction of the resin by exposure to lower the solubility in a developing solution to be described later. Therefore, the unexposed area 11 in the developing step is dissolved by the developer, and only the exposed region 10 remains on the surface of the substrate to become a mask. On the other hand, when the photosensitive resin film is formed of a positive photosensitive resin, the exposed region 10 is subjected to exposure to cut the resin, and the solubility in a developing solution to be described later is increased. Therefore, the exposed region 11 in the developing step is dissolved by the developer, and only the unexposed region 10 remains on the surface of the substrate to form a mask. [5] Developing Step Next, in the developing step, when the photosensitive resin film 9 is made of a negative photosensitive resin, the unexposed region 11 is dissolved by the developing solution, and only the exposed region 10 remains on the substrate. On the surface, it acts as a mask in the first etching step. On the other hand, when the photosensitive resin film 9 uses a positive photosensitive resin, only the exposed region 1 溶解 is dissolved by the developer, and the unexposed region 11 remains on the surface of the substrate, followed by the first It functions as a mask in the etching step. The developer used in the developing step can be used by a known person. Examples thereof include inorganic bases such as sodium hydroxide, potassium hydroxide, sodium carbonate, sodium citrate, sodium metasilicate, and aqueous ammonia; primary amines such as ethylamine 'n-propylamine; diethylamine and di-n-butyl Secondary amines such as butylamine, tertiary amines such as triethylamine and methyldiethylamine, alcoholamines such as dimethylethanolamine and triethanolamine, tetradecyl ammonium hydroxide, tetraethyl hydroxide An aqueous solution of a quaternary ammonium salt such as ammonium or trimethyl hydroxyethylammonium hydroxide; a cyclic aqueous amine such as 吼° and a slight bite; and an organic solvent such as toluene or toluene. The developing method in the developing step is not particularly limited, and methods such as dip development, spray development, brush development, and ultrasonic development may be employed. Fig. 57(d) is a view schematically showing a state in which development processing is performed on the photosensitive resin film 9 using a negative photosensitive resin. The unexposed area 11 in Fig. 57(c) is dissolved by the developer, and only the exposed region 10 remains on the surface of the substrate to become the mask 12. Fig. 57(e) is a view schematically showing a state in which development processing is performed on the photosensitive resin film 9 using a positive photosensitive resin. The exposed region 10 in Fig. 57(c) is dissolved by the developer, and only the unexposed region 11 remains on the surface of the substrate to become a mask 12 ° [6] first etching step, then In the etching step, the photosensitive resin film remaining on the surface of the substrate for a mold after the development step is used as a mask, and the substrate for the mold which is not covered is mainly etched, and the plated after the polishing is applied. The covering surface is formed with irregularities. Fig. 58 is a view schematically showing a preferred example of the latter half of the mold manufacturing method of the present invention. Fig. 58(a) is a view schematically showing a state in which the substrate 7 for a mold which is not covered by the mask portion 13 is mainly etched by the first etching step. Although the base material 7 for the mold at the lower portion of the mask 12 is not etched from the surface of the base material for the mold, it is engraved from the uncovered portion 13 as the etching progresses. Therefore, near the boundary between the mask 12 and the unmasked portion 13, the substrate 7 for the mold at the lower portion of the mask 12 is also etched. In the vicinity of the boundary between the mask 12 and the unmasked portion 13, the case where the mold substrate 7 under the mask 12 is also etched is referred to as side etching. Figure 59 is a schematic representation of 61 321900 201042295 showing the side touch. The broken line 14 of Fig. 59 shows the surface of the substrate for a mold which changes as the remainder progresses. The etching treatment in the first etching step is usually carried out by using a vaporized iron (FeCh) liquid, a vaporized copper (CuCl 2 ) liquid, an alkali etching solution (Cu(NH 3 ) 4 Cl 2 ), or the like, by causing corrosion of the metal surface, but also A strong acid such as citric acid or sulfuric acid may be used, or reverse electrolytic etching may be employed by applying a potential opposite to that at the time of electrolytic plating. The concave shape formed in the substrate for a mold when the etching treatment is applied differs depending on the type of the underlying metal, the type of the photosensitive resin film, the etching method, and the like, and therefore cannot be generalized, and when the etching amount is 1 〇 #m or less, The surface of the metal that is in contact with the engraving liquid is substantially equiaxed. The term "instant" as used herein refers to the thickness of a substrate that is cut by the remainder. The etching amount in the first scribe step is preferably from 1 to 50 Am. When the etching amount is not reached, the uneven shape is hardly formed on the metal surface, and the mold is formed into a substantially flat mold, so that the anti-glare property is not exhibited. In addition, when the etching amount exceeds 50 #m, there is a concern that the height difference of the uneven shape formed on the metal surface becomes large, and the image display device using the obtained mold to prevent glare is white turbid. phenomenon. The etching treatment in the i-th etching step may be performed by one etching treatment, or may be performed by etching twice or more. When the etching treatment is carried out in two or more steps, the total etching amount of the etching treatment of two or more times is preferably 丨 to 5 〇 #m. [7] Photosensitive resin film peeling step Next, in the photosensitive resin film peeling step, the remaining photosensitive resin film used as a mask in the second etching step is completely dissolved. In the photosensitive resin film peeling step, a peeling liquid is used to dissolve the photosensitive resin film of 62 321900 201042295. In the case of the peeling liquid, the same as the above-mentioned developing solution can be used. By changing the pH, the temperature, the concentration, the impregnation time, and the like, the green photosensitive resin film of the exposed portion is completely used by using the negative photosensitive property. The dissolution f removal j completely dissolves and removes the non-exposed portion when the positive photosensitive resin film is used. The method of removing the photosensitive resin film is not particularly limited, and methods such as dipping development, nozzle fog development, brush development, and super development may be employed.

〇w#Y. ^ Figure (b) shows the sinister display of the photosensitive resin film used as the mask 12 by the photosensitive resin film peeling step b, and the photosensitive resin film used as the mask 12 in the step. . By using the photosensitive resin film to form i=12 (10) engraved, the surface of the substrate for the mold is formed! The surface is concave [8] The second step of the cover is applied to the dream: the concave and convex surface (the first surface uneven shape 15) formed by the ft pair is in the shape of the concave and convex shape and the relaxation. Fig. 58(c) is a view showing a state in which the surface unevenness shape r5 formed by the fourth (4) process is more than = 16: thereby forming a state in which the unevenness ratio is the first in the outer path e-day k (the chrome-plated surface 17). Ze, hard silk, and ^ can be used to impart light to the flat or fresh surface. The friction coefficient is small, and the type is not particularly limited. The chrome of the key chrome or the like exhibits a good luster; the so-called gloss or decorative ore, which can be used for the plating tank. Chrome plating is usually carried out by electrolysis into an aqueous solution. Sa ft 3 ", 'Water chromic acid (Cr 〇 3) and a small amount of sulphur, the current density and electrolysis time can control the thickness of the ore chrome 321900 63 201042295. In the above-mentioned Japanese special opening 2〇〇2_1891〇 In the sixth publication, Japanese Patent No. 45472, and Japanese Laid-Open Patent Publication No. 2004-90187, the mineral chromium is used, but depending on the front layer of the mold and the chrome-plated shaft, the surface of the large ore is roughened or produced. As a result of the chrome plating, the result is 'transparent with a surface irregular shape obtained by the mold' containing an anti-glare film.) (4) Learning the direction of the material. The surface of the material is not suitable for the transparent substrate. The reason for the production of the anti-glare film is that the surface of the plating is polished after the chrome is removed, but it is not preferable to perform the surface polishing after the plating in the present invention. In the present invention, the underlying metal is subjected to wrought copper or nickel plating to solve the problem that is easily caused by chrome plating. Further, in the second plating step, it is not preferable to apply a forging other than the mineral. The reason is that 'in the mine cover other than chromium, due to hard Or the abrasion resistance is low, so the durability as a mold is lowered, and the unevenness is worn during use. The mold is damaged. It is difficult to obtain the anti-glare treatment using the mold and the anti-slip towel obtained from the mold. The possibility of sufficient anti-glare function is π, and the possibility of causing defects on a transparent substrate such as a transparent resin crucible is also high. Further, in the above-mentioned Japanese Patent Publication No. 2〇〇4_9〇187, The surface grinding after plating is also not preferable in the present invention. That is, it is preferably not provided in the step of the second ore cover step, and the chrome-plated surface is directly used as the transfer. The uneven surface of the mold on the transparent substrate. The reason is that 'there is a portion of the flat surface of the flat surface of the flat surface of the flat surface of the flat surface. Good shape control Γ read. =, in the manufacturing method of the mold of the present invention = the surface of the surface of the fine-grain surface is applied with a mineral layer of 10% depending on the type of the underlying metal, by the first 3 redness, degree, And coverage Type, thickness, etc.; the same, because: two two o 2 = slow: =, is _ degrees. If = = fruit and: full, and the concave and convex shape transfer two, i obtained on the transparent The optical properties of the substrate (5) glare film are not good. The other one is worse, and it will be too thick, and the productivity will be trapped. Therefore, it is a protrusion of nodule. Within the range of plating defects: =:, the thickness of the secret is preferably 1 to 1 〇 - and live in the range of the horse to the β. 0 degree in the mirror layer of = is better to form the Vickers hard reduction Wei Tit durable __'μ hardness in forging, etc., abnormality, plating bath composition during plating treatment, high electrolysis conditions, high; In the above-mentioned (7) photosensitive resin film peeling step, it is preferable that the uneven surface formed by the etching step between the steps of the second step by the retort processing becomes gentle and gentle 321900 65 201042295 The second etching step. In the second etching step, the first surface uneven shape 15 formed by the first etching step using the photosensitive resin film as a mask is made gentle by etching. By the second etching treatment, the portion of the first surface uneven shape 15 formed by the first etching treatment is steeply inclined, and the anti-glare film or the like produced by the obtained mold is transparently provided with anti-glare treatment. The optical properties of the substrate will change in a good direction. Fig. 60 is a view schematically showing a second surface unevenness having a gentle surface inclination by making the first surface uneven shape 15 of the mold substrate 7 gentle and gentle, and making the surface inclined steeply gentle by the two etching steps. The state of shape 18. In the etching process in the second etching step, similarly to the first etching step, a ferric chloride (FeCl3) solution, a vaporized copper (CuCh) solution, a test engraving solution (Cu(NH3)4Cl2), or the like is usually used. The metal surface is rotted, but a strong acid such as citric acid or sulfuric acid may be used, or reverse electrolytic etching may be employed by applying a potential opposite to that at the time of electrolytic plating. The degree of relief of the concavities and convexities after the application of the etching treatment differs depending on the type of the underlying metal, the button indentation, and the size and depth of the concavities obtained by the first remaining step, and therefore cannot be generalized, and the maximum factor in controlling the degree of relaxation It is the name of the surname. The amount of etching referred to herein is also the thickness of the substrate which is cut by etching, similarly to the first plating step. When the amount of etching is small, the effect of reducing the surface shape of the unevenness obtained by the first plating step is not sufficient, and the uneven shape is transferred to a transparent substrate such as a transparent film to obtain an anti-glare portion. The optical properties of the transparent substrate (anti-glare film) are not good. On the other hand, if the etching amount is too large, the uneven shape is almost eliminated, and the mold is substantially flat, so that the anti-glare property does not appear. Therefore, the amount of the etch is preferably in the range of 1 to 5 0 /z m 66 321900 201042295 in the range of -4 to 2 G/zm. In the same manner as the first etching step, the 余2 retrace processing may be performed by 丄&amp; When := is 2 or more, more than 2 times or more is preferably 1 to 50 #m. 〇Τ 〇Τ The anti-glare film obtained by the anti-glare treatment method of the present invention and the anti-glare film obtained by the method for producing an anti-glare film are formed by controlling the fine unevenness of the transparent substrate and the precision of the anti-glare treatment. Anti-glare ν. The phenomenon of white turbidity occurs in the image display device and does not produce flashing, (4) high contrast. The present invention is not limited to the following examples, but the present invention is not limited to the following examples, and the present invention will be described in more detail, and is limited to the above embodiments. &lt;Examples 1 to 3 and Comparative Examples 1 to 2&gt; An inch roll having a diameter of 200 mm was prepared (a ballad plating was applied according to JIS Α5〇5. The copper μ special surface was made of a copper plating layer) / Thin silver plating layer / surface copper plating layer. The thickness of the plating layer is set to about 200 / zm. The surface of the copper plating surface is subjected to a "surface grinding" coating a negative photosensitive resin Polishing the surface of the copper ore and drying it to form a photosensitive resin film. B. Next, the five types of patterns shown below are rubbed by the laser light to expose v on the photosensitive resin film. Development. Exposure and development with laser light utilizes laser stream FX (Think

Laboratory (stock) system to carry out. 67 3219〇〇 201042295 y) Pattern I (Embodiment 1): A pattern of a part of the unit® shown in Fig. 61 is displayed in reverse. The unit pattern is a 32.768 mm square pattern produced by a resolution of 12 dpi. Figure 61 shows the square of 1·=4ιηιη. The unit pattern shown in Fig. 61 is such that the average difficulty is 16/zm at the density of 2_hearts 2, and is not distributed as shown in Fig. 62. The pattern 'Using the spatial frequency range lower limit B of i times is G.〇4 (^mi, the upper limit of the spatial frequency range τ is 〇, so '2χ(τ_Β)/(10))=G. 55), and penetrates The peak value has a band-pass (four) device with a steeply inclined asymmetric shape on the spatial frequency side, and then the obtained second instance is binarized by the threshold method. The lower limit value B of the spatial frequency range of the obtained unit pattern is, and the upper limit value T of the spatial frequency range is 0.067 / ζ π Γ 1. (2) Pattern Π (Embodiment 2): A pattern in which a unit pattern of a trowel is displayed in Fig. 63 is repeatedly arranged. The unit pattern is a 32.768 mm square pattern produced by a resolution of 128 (10) dpi. The 帛63 image is cut out of the 1.024 coffee square. The unit pattern shown in Fig. 63 is used for the i-th pattern shown in Fig. 62, and is used for i-times with the above-mentioned pattern, and then the second method is binarized. Then, use the same bandpass == 9 times. The lower limit value B of the spatial frequency range of the obtained unit pattern is 〇.〇〇# m 1 ' The upper limit value τ of the spatial frequency range is 〇. ^ (3) Pattern Ι π (Embodiment 3): A pattern of a unit pattern showing a portion in Fig. 64 is repeatedly arranged. The unit pattern is a 32 768 mm square pattern produced by a resolution of 128 〇〇 dPi. Figure 64 is a cut out of the 1900 mm square of 321900 68 201042295 &lt;. The unit pattern shown in Fig. 64 is the same as the first pattern shown in Fig. 62, and the same bandpass filter as that of the user of the pattern I is applied once, and then the obtained second pattern is subjected to the threshold method. After binarization, the same passband filter is used 19 times. The upper limit value B of the spatial frequency range of the obtained unit pattern is 0. 047/i m_1, and the upper limit value T of the spatial frequency range is 0. 067/ζπΓ1. (4) Pattern IV (Comparative Example 1): A pattern in which a part of the unit pattern is displayed in Fig. 65 is repeatedly arranged. The unit pattern is a pattern of 20.944 mm squares produced by resolution of 12800 dpi. Fig. 65 shows the 1.024 mm square cut out. The unit pattern shown in Fig. 65 was produced by irregularly distributing dots having an average spot diameter of 16//m at a density of 1419/mm2. (5) Pattern V (Comparative Example 2): A pattern in which a part of the unit pattern is displayed in Fig. 66 is repeatedly arranged. The unit pattern is a pattern of 20.944 mm squares produced at a resolution of 12,800 dpi. Figure 66 shows the Q 1. 024mm square. The unit pattern shown in Fig. 66 was produced by irregularly distributing dots having an average spot diameter of 16//m at a density of 1419 / mm 2 . The five patterns I to V described above are simultaneously exposed/developed on the photosensitive resin film by laser light, and then the first etching treatment is performed with a copper chloride solution. The etching amount at this time was set to 3 #m. The photosensitive resin film was removed from the roll after the first etching treatment, and the second etching treatment was again performed with a copper chloride solution. The etching amount at this time was set to 10 / / m. Then, chrome processing is performed to make a mold. At this time, the thickness of the chrome plating was set to 4/z m. 69 321900 201042295

Acetate ethyl iodide photohardenable resin composition (3) Na IC 8〇6T (manufactured by Otsuka Ink Chemical Industry Co., Ltd.) as a solution of 50% by weight concentration and then Lucirin tpq belonging to photopolymerizing silk agent (The basf company, chemical name: 2'4,6-methylfgi-based diphenylphosphine oxide) was added in an amount of 5 parts by weight per 10 parts by weight of the curable resin component to prepare a coating liquid. This coating liquid was applied onto a film of cellulose triacetate (10) having a thickness of 8 Mm so as to have a coating thickness after drying of 1 G/m, and was allowed to make a money for 3 minutes in a dryer set to 赃. The rubber roller reduces the film behind the yarn to the uneven surface of the previously obtained mold and makes it adhere to each other, and the photocurable resin composition layer becomes the mold side. In this state, the high-pressure mercury (4) light from the strong leak/(10) 2 is irradiated from the TAC film side, and the amount of light converted into the h-line is 20 〇 mJ/cm 2 to cure the photocurable resin composition layer. Then, the tac film and the cured resin are peeled off from the mold to form a transparent antiglare film comprising a laminate of a cured resin having a concavity and convexity on the surface and a TM film, and having five uneven surface shapes corresponding to the patterns I to V. &lt;Example 4&gt; In the same manner as in Example 1 except that the unit pattern of a part shown in Fig. 67 was repeatedly exposed and developed on the photosensitive resin film by one roll by laser light; An anti-glare film was produced in the same manner as in Example j except that it was obtained. The same operation was performed twice to obtain a total of 2 anti-glare films. The unit pattern shown in Fig. 67 is a pattern of 32.768 mm square generated by the resolution of l2_dpi, and Fig. 67 is a figure of one square 24_ square. The unit pattern in Fig. 67 is the second pattern after the bandpass filter 321900 70 201042295 chopper is used to create the second pattern. The hunting is binarized by the error diffusion method to create the third pattern, and then repeated 60 times. The fourth pattern made by Monte Carlo. The first pattern used is a bit map of 32 bits of 32.768 mm square generated by a resolution of 12800 dpi, and PIXCEL[x,y] is substituted for a quadratic element having an 8-bit depth, and substituted into PlXCEL[x , y] = R|&gt; + y]xlmage Width]x255 and the producer. x and y are the coordinates of the pixels in the image, and Image Width is the pixel width of the X coordinate. The arrangement of R[] is based on the selection of the Random Double method of the Random Level method contained in the ".NET Framework2. Library". The value between 〇 and 1.0 is subtracted by the Knuth random number generator. The quasi-random sequence generated by the algorithm. For the band pass-through device, the lower limit value B of the spatial frequency range is 0. 045 /inf1, and the upper limit value τ of the spatial frequency range is 〇. 080 β Therefore, 2x(TB)/(T+B) = 0. 56), and the through-frequency domain peak has a band-pass filter of asymmetrical shape with a steeper slope on the low spatial frequency side. Moreover, in the case of the error diffusion matrix, the error diffusion matrix having the diffusion distance of 3 shown in FIG. 36 and the diffusion distance shown in FIG. 37 are 4 in the ratio of 0.4:0.6. The diffusion matrix is combined (Fig. 36 x0. 4+ Fig. 37 χ〇 6). The spatial frequency range lower limit value Β of the unit pattern shown in Fig. 67 is 0.045 / ζ π Γ 1 '. The spatial frequency range upper limit value Τ is 0.086 / / nf1. Fig. 68 is a view showing the spatial frequency distribution of the unit pattern used in the first to third embodiments. Fig. 69 is a view showing the spatial frequency distribution of the unit patterns used in Comparative Examples 1 to 2. Fig. 70 is a view showing the spatial frequency distribution of the unit pattern used in the fourth embodiment. &lt;Example 5&gt; 71 321900 201042295 A person who applied a ballad plating to a surface of an aluminum roll having a diameter of 200 nm (according to JIS A5056) was prepared. The copper Barrat plating is composed of a copper plating layer/thin silver plating layer/surface copper plating layer, and the thickness of the entire plating layer is set to be about 200//m. The copper-plated surface was mirror-finished by applying a positive photosensitive resin to the polished copper-plated surface and drying it to form a photosensitive resin film. Then, the unit pattern of a portion shown in Fig. 71 is repeatedly exposed/developed on the photosensitive resin film by laser light. The exposure and development by laser light were carried out by using Laser Stream FX (Think Laboratory). The unit pattern shown in Fig. 71 is a 32.768 mm square pattern produced by a resolution of 12,800 dpi. Figure 71 is a cut out of 1. 024mm square. In the unit pattern shown in Fig. 71, the second pattern is created by using a band pass tear wave for the first pattern, and then the third pattern is binarized by the error diffusion method, and the Monte Carlo method is applied 60 times. The fourth pattern produced. The first pattern used is a bitmap of 32. 768mm square octet generated by a resolution of 12800 dpi, and for a quadratic arrangement PIXCEL[x, y] having an 8-bit depth, substituting piXCEL[x , y] = R[x + yxlmage Width]x255 and the producer. Where x and y are the coordinates of the pixels in the image, and Image Width is the pixel width of the χ coordinate. Arrangement R[] is a subtraction algorithm from Knuth's chaotic generator generated by selecting the value of 〇. 〇 and 1 由 generated by the Random Level Next Double method contained in the ".NET Framework 2. 〇 Level Library". The resulting pseudo-random sequence. For the bandpass filter, the lower limit of the spatial frequency range is 321900 72 201042295 Β is 0. 055/^-1, the spatial frequency is this, 2χ(Τ-Β)/(Τ+Β) = 〇. 58广^ Τ is 〇·100/ζπΓ1 (Inss function-type band-pass filter ^The shape of the peak of the transmission frequency domain is high, and the ratio of 0.9:0.1 is used for the third and the difference diffusion matrix is the error diffusion. The matrix and the 38th figure - the diffusion distance of 4 arrays are combined (the 37th _ ; ^ diffusion distance is 5, the spatial diffusion of the unit pattern of the error diffusion moment is shown in Figure 71. The upper limit of the frequency range Τ is about 〇1 is about 〇· 〇55/zm 1 'The space of the unit pattern shown in the empty figure ^ 1GG/am ° Figure 72 shows the 71st, _ rate distribution map. Then, The system was set to 5#m with a copper chloride solution. The etched amount of the film was again treated with a copper chloride solution; the roller removal photosensitive resin was set to 8, and then, = Processing. At this time, the thickness of the mineral deposit is set to 4/r chrome plus f. At this time, the acetic acid g~ lysis hardenable resin composition grandIc Na ^ Japan Water Chemical Industry Co., Ltd., as a solution of 5G wt% concentration, and will be a photopolymerization initiator Lueirin (manufactured by Hall F, = subname. 2, 4, 6-dimethyl The benzhydryldiphenylphosphine oxide was added in an amount of 5 parts by weight per 2 m of the curable resin component to prepare a coating liquid. The liquid was applied to a cellulose acetate (TAC) film having a thickness of 80 Å. After drying, the coating thickness is 1 〇 _, and the drying machine is set to dry 3 胄. The dried film is pressed by the rubber roller to the concave and convex surface of the previous Z mold, and is closely adhered. In addition, the photocurable resin composition θ is set to the mold side. In this state, the light from the high-pressure mercury lamp having a strength of 20 mW/cm 2 is irradiated from the tac film side so that the amount of light converted into h lines is 321900 73 201042295 200 mJ/cm. The photocurable resin composition layer is cured, and then the TAC film is peeled off from the mold together with the cured resin to form a transparent antiglare film composed of a laminate of a cured resin having a concavity and convexity on the surface and a TAC film. Examples 1 to 5 and Comparative Examples 1-2 The glare film was subjected to the evaluation test described below. (1) Scintillation Evaluation The butterfly system was evaluated by the following method. b First, the pattern of the unit cell 60 shown in the plan view of Fig. 73 (a) was prepared to be about 40 π π χ χ 25 A reticle in which the range of the coffee is regularly arranged. In the unit cell 60, a key-shaped chrome-shielding pattern 61 is formed on the transparent substrate by a line width 10, and a portion where the chrome-shielding pattern 61 is not formed is an opening. 62. The mask is given "nominal S1Ze" (unit: PPKpixel per inch) according to the size of the unit cell. For example, the unit cell length unit unit width of the mask having a resolution of 90 ppi is 2, #mx94#m, and the opening length of the opening X is 272ymx84ym. The unit cell was fabricated according to the numerical values of Table 1, and a mask of the total pattern was prepared in the range of the resolution nominal size to 180 ppi. 321900 74 201042295 Table 1 Resolution nominal size (ppi) Unit cell opening size (// m) Wide size (/W m) Long size (// m) Wide size (# Π1) 50 508 169 498 159 60 423 141 413 131 70 362 120 352 110 80 317 105 307 95 90 282 94 272 84 100 254 84 244 74 120 211 70 201 60 140 181 60 171 50 160 158 52 148 42 180 141 47 131 37 . Next, as in Figure 73 (b), the chrome-shielding pattern 61 of the reticle 63 is placed upwards and placed in the light box 65 (the light source 66 is disposed in the light box), and the glass plate 67 of the thickness of 1.1 is 20/m thick. The sample in which the anti-glare film 70 贴 was adhered to the adhesive was placed on the reticle 63, and visually observed from a position (visual observation place 69) of about 30 cm from the sample, whereby the presence or absence of scintillation was evaluated. This evaluation was performed for each of the prepared masks having nominal sizes of different resolutions. In the above evaluation, depending on the characteristics of the anti-glare film, flicker was observed in a reticle having a resolution higher than the nominal size. At this time, the nominal size of the resolution is used to evaluate the flash. Specifically, a method of discriminating the evaluation value will be exemplified. First, when performing the functional evaluation, strong flicker was observed in the mask with a resolution of 90 ppi, and in the 75 321900 201042295 mask with a resolution of 80 ppi, when the flicker was observed, 8 was given. However, depending on the characteristics of the anti-glare film, only a weak flicker state is observed in the resolution nominal size cover. In order to distinguish the situation from ===production_scissing -==resolution nominal size 8G_(4) 85pρι of the intermediate value of PPi is used as the scintillation evaluation to distinguish the above two states. (2) Evaluation of the penetration characteristics The haze of the anti-glare film was measured using a haze meter (Haze meter, Murakami Color Technology Research Institute 9). The results of the above evaluation test, the production method of the unit pattern, and the production conditions of the mold are shown in Table 2. Further, in Example 4, the evaluation results of the two anti-glare films are respectively shown. Table 2 Example 1 Capital Example 2 Example 3 Comparative Example 1 Comparative Example 2 Example 4 Example 5 Method for making a unit round case Using a y 1 filter, whether or not there is a binarization method 〇——————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————————— Amount (β m) 10 10 10 10 10 10 8 Chrome plating thickness (β m) 4 4 4 4 4 4 4 Glare evaluation (ppi) 80 80 85 50 below 65 110 130 95 Haze (9〇0.6 0. 7 0. 7 1.5 1. 1 0.7 0.8 0. 4 76 321900 201042295 By the scintillation evaluation test of the reticle, according to the example of the use of the band-pass low spatial frequency component reduction pattern, the film i, and the basis Compared with the anti-glare to the anti-glare of the ith pattern of the regular distribution, it is confirmed that the upper limit of the former does not produce a higher level and the performance is better, and the error expansion is used as the second-order material. The fourth material is produced by the two anti-glare films of ^ = and the anti-glare of Example 5, and the threshold method is used. Compared with the anti-glare films of Examples 1 to 3, the former did not observe flashing even in a higher first cover, and exhibited a better optical property. <Example θ> Prepared for an aluminum roll having a diameter of 200 (10) (according to Surface of JIS Α5〇56) • Bronze ballad plating. Copper Barat plating is made of copper plating/thin silver plating/surface copper plating. The thickness of the entire coating layer was set to about 200 #m. The copper plating surface was mirror-polished, and a positive photosensitive resin was applied onto the surface of the polished copper and dried and dried to form a photosensitive resin film. Then, the unit pattern of a portion shown in Fig. 74 is repeatedly exposed/developed on the photosensitive resin film by laser light. The exposure and development by laser light is performed by laser stream FX (Think Laboratory) The unit pattern shown in Fig. 74 is a 32. 768mm square pattern produced by a resolution of 12800 dpi. The 74th figure is a cut out of 1. 024 mm square. The unit pattern shown in Fig. 74 Apply bandpass filtering relative to the first pattern After the second pattern was created, the third pattern of the binarization 77 321900 201042295 was created by the error diffusion method, and the fourth pattern produced by the Monte Carlo method was used repeatedly. The first pattern used was 1000 〇. The density of /mm2 makes the average point diameter of 8/im points irregularly distributed to the producer. At this time, 'because the point is evenly distributed, the triangular lattice corresponding to the set point density is set' so that each of the center coordinates X and γ of the point is from the lattice point to the lattice of the set triangular lattice. Displacement, thereby creating a pattern. Further, the decision of the coordinates after the displacement is a program code generated by C#() shown below (a programming language developed by Microsoft Corporation, and the language specification is defined by "JISX 3015 programming language c#"). The point position is irregularly displaced by giving the coordinate value (X or γ) of the function as the grid point of the Average displacement and Devia1; i〇ri to 〇.3x15//in'. At this time, the "RandomFunction()" in the pseudo-number #code is given as a seed by giving the value 607 to the SIMD oriented Fast Mersenne Twister program 'SFMT verl· 3. 3 installed by the Hiroshima University team. (The code generated by C# used in Embodiment 6) //cx,cy : Displays the X coordinate and γ coordinate of the center of the newly drawn point. &quot;Px ’ py : Displays the X coordinate and γ coordinate of the set triangle lattice point. //pD : 0. 3 //CoreSize : the diameter of the point cX = NormalRandom(px, pD * CoreSize); cY = NormalRandom(py, pD * CoreSize); // The normalization function of random numbers // RandomFunction(): The function that returns the random number 0 &quot;RandomFunctionValueMaxO: A function that returns the maximum value of the random number 321900 78 201042295. // Math : .NET Framework Math Library * public double Norma1Random(doub1e Average, double Deviation) { double buff = 0 ; buff = Deviation*Math. SqrtC-2 * Math.Log (((double)RandomFunction()/( Do uble) 〇

RandomFunctionValueMaxO))) * Math. Sin(2 * Math. PI * ((double)Rando mFunction() / (double)

RandomFunct i onVa1ueMax()))+Average ; if(buff&lt;0){buff=0 ; }; return buff ; } For the Qualcomm filter, the lower limit B of the spatial frequency range is 〇〇. High pass filter. Moreover, in the case of the error diffusion matrix, the error diffusion matrix having a diffusion distance of 4 as shown in FIG. 37 and the error diffusion of the diffusion distance of 5 shown in FIG. 38 are used in a ratio of 0.9.0.1. The matrix is combined (Fig. 37, X0.9+38, Fig. 丨). The lower limit of the spatial frequency range 单位 of the unit pattern shown in Fig. 74 is approximately. Then, the first etching treatment is performed with a copper chloride solution. The etching amount at this time is set to . The photosensitive resin film was lightly removed by the first # etching treatment, and the second remaining treatment was performed again with the copper chloride solution. At this time, the amount of the balance is set to 18# and then the mineral processing is performed to make a mold. At this time, 321900 79 201042295 chrome thickness is set to 4.乂 乂 ga ga / / / 乂 GRA GRA GRA GRA GRA GRA GRA GRA GRA GRA GRA GRA GRA GRA GRA GRA GRA GRA GRA 大 大 大 大 大 大 大 大 大 大 大 大 大 大 大 大 大 大 大 大 大 大 大 大 大The company name, chemical name: 2, 4, 6-trimethylsulfonyldiphenylphosphine oxide, was added in an amount of 5 parts by weight per 1 part by weight of the hardenable component to prepare a coating liquid. This coating liquid was applied onto a cellulose triacetate (tac) film having a thickness of 8 G/a m so that the coating thickness after drying was 1 Q/m, and it was dried in a dryer set to a step for 3 minutes. The dried film was pressed against the uneven surface of the previously obtained mold with a rubber and adhered thereto, and the photocurable resin composition layer was formed on the mold side. In this state, the light from the TAC film side is irradiated with light from a TAC film side to a compacted mercury lamp having a strength of 2 〇 mW/cm 2 so that the light curable resin composition layer is cured, so that the amount of light converted by the h line is 200 mJ/cm 2 . Then, the TAC film and the cured resin were peeled off from the mold to form a transparent glare film composed of a laminate of a cured resin having a concavity and convexity on the surface and a TAC film. A pattern in which a unit pattern of a part shown in FIG. 75 was repeatedly arranged in the same manner as in Example 6 was produced except that the threshold method was used for the same, and a mold was produced in the same manner as in Example 6 except that the pattern was used. Obtain an anti-glare film. Fig. 76 is a graph comparing the spatial frequency distribution of the pattern shown in Fig. 74 with the spatial frequency distribution of the pattern shown in Fig. 75. As seen from Fig. 76, in the pattern of Fig. 74 using the error diffusion method, the low spatial frequency component is further reduced. 321900 80 201042295 &lt;Example 8&gt; ^ A mold was produced in the same manner as in Example 6 except that the pattern of the unit pattern shown in Fig. 77 was repeatedly arranged to obtain an anti-glare film. The fourth pattern shown in Fig. 77 is a 32.768 mm square pattern produced by a resolution of 12800 dpi, and the 77th figure is one of 24 square squares cut out. In the fourth pattern, the second pattern is binarized by the error diffusion method of the error diffusion matrix shown in FIG. 37 according to the error diffusion distance of 4 for the first pattern, and the third pattern is created, and the second pattern is created. The maker of the second Carlo system 60 uses the spatial frequency range lower limit value B and the spatial frequency range upper limit value τ by the above equations (I) and (u) [MainPeriod=12 (/zm), Bandwidth=20(%)], and the pass-through frequency domain peak shape is a Gaussian bandpass filter. The first pattern is a 32. 768mm square 8-bit bitmap image generated by a resolution of 12800 dpi and for a quadratic arrangement PIXCEL[x, y] having an 8-bit depth, substituting PlxcEL[x, y] = R[x+yxImageWidth] ❹x255 and the producer. Where x and y are the coordinates of the pixels in the image, and Image Width is the pixel width of the X coordinate. Arrangement R[] is based on the value of the nu. io and io generated by the Random Level Next Double method contained in the ".NET Framework 2.0 level library". The Knuth 籥L number generator subtraction algorithm is used. The resulting pseudo-random sequence. &lt;Example 9&gt; A pattern of a unit pattern of a part shown in Fig. 78 was repeatedly arranged in the same manner as in Example 8 except that the threshold method was used for binarization. Then, a mold was produced in the same manner as in Example 8 except that this pattern was used, and 81 321900 201042295 was obtained to obtain an anti-glare film. Fig. 79 is a view comparing the spatial frequency distribution of the pattern shown in Fig. 77 with the spatial frequency distribution of the pattern shown in Fig. 78. As is apparent from the figure (7), in the pattern of the 77th drawing of the error diffusion method, the anti-glare treatment such as the anti-glare film produced by the method of the present invention is used for the anti-glare treatment. The shape of the fine uneven surface reflecting the pattern having a small low spatial frequency component is such that no flicker is generated, and sufficient anti-glare property is exhibited, and a white turbidity phenomenon is generated. In addition, since the haze is also low, there is no reduction in contrast when disposed on the shirt image display device. In addition, since it is difficult to prevent the material from being isolated, it is also suitable for the engraving. <Example: Using the band pass; patterning and evaluation by the wave filter> The pattern 1 is created by the method shown below. 15. 24^Γ 1 : For the density of 1111 / coffee 2 to make the average point diameter / point shame made in the 80th towel shows the _ part of the pass, the second use of the spatial frequency range lower limit Β, is 0. After the high-pass chopper of the heart is mounted as the second pattern, the pattern 1 is obtained by binarizing m by the value between the values. Figure _. In addition, in the above-mentioned first drawing "production: the first pattern of the pattern 1 is partially enlarged (four) is the same as two = ^ (2) pattern 2 · · for the homogenization of the knife cloth for the map /, t point. The ratio of ... will be the third pattern of the 3rd:: error diffusion matrix and the diffusion distance shown in Fig. 38 == 321900 82 201042295

The error diffusion matrix synthesized by the matrix (the error diffusion method of 371. 9+ 38th diagram 〇. D, obtaining the pattern belonging to the third pattern, Fig. 82 shows a partial enlargement of the pattern 2. (3) Pattern 3: Apply Monte Carlo method to pattern 2 repeatedly to obtain pattern 3 belonging to the fourth pattern. Fig. 83 shows a partial enlargement of pattern 3. (4) Pattern 4·except for use The pattern 4 was obtained in the same manner as the pattern 1 except that the first pattern B was partially formed by an irregular distribution of the average 〇 point control at a density of 16 //mm 2 . A partial enlargement of the pattern 4 is displayed. (5) Pattern 5: For the second pattern used for the pattern 4, the diffusion distance shown in Fig. 37 is used in a ratio of 〇.9.0.1. The error diffusion method of the error diffusion matrix of 4 and the error diffusion matrix of the diffusion distance of 5 shown in Fig. 38 is synthesized by the error diffusion method (Fig. 37 x 0. 9 + Fig. 38 χ〇. 1). 3 pattern of pattern 5. Figure 86 shows the pattern of partial enlargement of pattern 5. (6) Pattern 6: Monte Carlo method 60 times is applied to the pattern 5 to obtain the pattern 6 belonging to the fourth pattern. Fig. 87 shows a partial enlargement of the pattern 6. (7) Pattern 7: except for using a density of 2,500 / Å 2 The pattern 7 is obtained in the same manner as the pattern 1 except that the dot pattern is an irregular distribution of dots of 16 #m, and a part of the first pattern C is displayed in Fig. 88. Fig. 89 shows that the pattern 7 is partially enlarged. Fig. 8 (8) Pattern 8: For the second pattern used for the pattern 7 production, the system 83 831900 201042295 uses the error of the diffusion distance shown in Fig. 37 to be 4 in the ratio of 0·9:0.1. The error diffusion method of the diffusion diffusion matrix (Fig. 37 x. 9+ 38th figure x 〇. 1) synthesized by the diffusion matrix and the error diffusion matrix of the diffusion distance of 5 shown in Fig. 38 is obtained by the third pattern. Pattern 8. Fig. 90 shows a partial enlargement of the pattern 8. (9) Pattern 9: The Monte Carlo method is applied 60 times over the pattern 8 to obtain a pattern 9 belonging to the fourth pattern. A partial enlargement of the pattern 9 is shown. (10) The pattern 10 is used except that the average dot diameter is 1 by a density of 4444/mm2. A pattern 1Q is obtained in the same manner as the pattern!, except that the second pattern D is partially distributed and displayed in Fig. 92. Fig. 93 shows a partial enlargement of the pattern 1〇. Π) Pattern 11: For the second pattern used for the pattern 10, an error diffusion matrix having a diffusion distance of 4 shown in Fig. 37 and a ratio shown in Fig. 38 are used at a ratio of 0.9: 〇·1. The error diffusion method synthesized by the error diffusion matrix with a diffusion distance of 5 (the first figure (9) is the error diffusion method of Fig. 38), and the pattern 属于 belonging to the third pattern is obtained. Fig. 94 shows a partial enlargement of the pattern 11. (12) Pattern 12: The Monte Carlo method is applied over the pattern u for 6 times to obtain the pattern 12 belonging to the fourth pattern. Fig. 95 is a view showing a partial enlargement of the pattern 12. (13) Pattern 13: A part of the first pattern E is displayed in Fig. 96, except that the first pattern E is displayed in the 96th figure by using an irregular distribution of dots having an average dot diameter of 8/ζιη at a density of ι0000/mm2. The pattern i similarly obtains the pattern 13. 321900 84 201042295 Figure 97 shows a partial enlargement of the pattern 13. Sun, heavy (Tian 14) pattern 14: For the second figure used for the production of pattern 13, it is for the use of 〇9. η, Ζ Ζ pattern, ., 0·1 ratio to expand the picture shown in Figure 37 The error diffusion matrix is the first jfj $ _ &gt; n , and the diffusion matrix is integrated into:: Figure 38 is the distance of 5 error η ^ The error diffusion matrix (Figure 37) The error expansion method of the stomach ο1) obtains the f9fHx pattern belonging to the third pattern and displays a circle in which the pattern 14 is partially enlarged. Pattern...第98Ο Ο

(15) Pattern 15: The Monte Carlo method is used to obtain the map macro ις belonging to the 4th figure. “N people are used in the pattern 14, and the partial magnification is shown. The figure display shows the pattern U. 】 _ to the second, the second two = the frequency distribution. The first ^ 1 Λ 糸 shows the spatial frequency distribution of the pattern 1 to 15. This 'dire 106 picture of the sac sac plaque pattern: the degree of reduction of the tank, 1 〇 6 _ =: : When the different first-first pattern is used, it can be used by the application of the high-level device, the error diffusion method, the Monte Carlo method, and the operation of the m-caro method. The efficiency reduction complex II is particularly effective in the third pattern using the error diffusion method. The fourth pattern, when the low spatial frequency component is reduced, is different from the bandpass chopper, because it is not in production. 2: ―5 疋 疋 upper limit 'Therefore there will be doubts about the occurrence of isolated points to 15 as shown, if the first pattern used is to be clicked, then as shown in Figure 1〇7, there is no majority The situation in which isolated points are generated. 321900 85 201042295 On the other hand 'in use, it will be shown on the 108th

The first pattern is the same as the creation of the display device 2 by the upper device and the threshold value method. The 109th display shows the same method for the production of the first pattern 1 shown in the first FIG. 8 and the high-pass filter line is applied. The pattern obtained by binarization is partially enlarged.

The method of +丄^ is the same. The method is to use a high-pass test and a partial enlargement of the pattern obtained by the error method. Fig. 111 is a view showing a pattern obtained by using a high-pass filter, a binarization by error diffusion method, and a clock method by the same method as the above-described pattern 3 in the same manner as the above-described pattern 3. A partially enlarged view. Fig. 112 is a diagram showing the isolated occupational records of the patterns shown in Figs. (10) to (1). Figure 113 is a comparison of the spatial frequency distribution of the graph (10) shown in Figures (10) to (1). From Fig. 112 and (1), even if the case contains many high-altitude components, by using the Qualcomm and Monte Carlo methods, the low-frequency components can be sufficiently reduced. The isolated point produces fewer good patterns. Fig. 1 is an enlarged view showing a preferred example of a first pattern which can be produced by a plurality of points in the anti-glare treatment method and the anti-glare manufacturing method of the transparent substrate of the present invention. 321900 86 201042295 Fig. 2 is a view showing a preferred aspect of the first pattern which can be used for the antiglare treatment method and the method for producing an antiglare film of the transparent substrate of the present invention, wherein the raster image is determined by a random number. An example of a picture. • The 帛3 image shows a diagram in which the - part of the first pattern shown in Fig. 2 is enlarged. Fig. 4 is a comparison of the quadratic array obtained by the second case (irregular dot pattern) which is produced by irregularly arranging a plurality of points, by the high-speed Fourier transform (FFO converted to the (four) frequency domain (four) frequency distribution A random number determines the image of the spatial frequency distribution obtained by the frequency domain in the frequency domain of the raster image (random raster image). The figure 5 shows the second layout of the sub-element of the i-th image, which is converted into the space brother = the obtained piece. The first picture of the 空间 空间 空间 空间 空间 空间 〇 〇 〇 〇 〇 〇 〇 〇 〇 〇 〇 〇 〇 The result-example diagram. The industry, ▽, and inter-work frequency distribution diagram 7 shows the spatial frequency domain extracted by using Qualcomm (penetration, the shape of the penetration belt in the middle). The peak shape of the penetrating zone in the transport (penetration zone) is based on the spatial frequency domain (penetration zone) (four) Space km, mountain-one inverse bandpass filter frequency 3219〇〇87 201042295 Figure 11 shows Another example of the shape of the penetrating band peak in the band-passing zone (penetration zone) is used. The inter-work frequency 12® system shows the peak of the penetration band through the transport (four) wave Another example of the shape of the inter-work frequency domain (an example of the peak shape of the penetration band of the enthalpy). The 1st picture of the 1st frequency is shown by using the band-passing field (penetration band) Another example of the penetration of the peak shape is shown in Fig. 15. The first pattern of the spatial frequency distribution shown in Fig. 5 is shown in Fig. 2 with the shift of the occupational wave rate: Fig. 16 is an enlarged view showing an example of a second pattern made by applying a band pass filter to the second pattern shown in Fig. 1. Fig. 17 shows the value of 2χ(τ_Β)/(τ + Β), A graph showing the relationship between the maximum value of the autocorrelation coefficient obtained by binarizing the second round obtained by using the π-pass filter by the value method. Fig. 18 shows the relationship of 2χ(τ_Β)/(τ + Β) The value and the threshold value method are used to calculate the number of isolated small dots of the pattern obtained by binarizing the second pattern obtained by the band pass filter. Figure 19 shows the distribution of the gray scale indicators obtained by analyzing the frequency map of the gray scale index for the image data shown in Fig. 16. Fig. 20 shows the second pattern binarized by the threshold method. An enlarged view of one example. Fig. 21 shows the spatial frequency obtained by converting the second pattern of the binarized second pattern 88 321900 201042295 shown in Fig. 2 by the fast Fourier transform (FFT) into the spatial frequency domain. The map of the distribution - Figure 22 is used to illustrate the diffusion weight map of the conversion error in the generally known error diffusion matrix. Figure 23 shows the error diffusion method based on the matrix of Floyd &amp; Steinberg. A case of an example of the third pattern is partially enlarged. Fig. 24 is a view showing a case where a third pattern obtained by using the error diffusion method of the matrix of Jarvis, Judis and Nink is partially released. Fig. 25 is a partially enlarged view showing an example of a third pattern obtained by using the error diffusion method of the matrix of Stucki. Fig. 26 is a partially enlarged view showing an example of a third pattern obtained by using the error diffusion method of the matrix of Sierra 3 Line. Fig. 27 is a partially enlarged view showing an example of a third pattern obtained by the error diffusion method based on the matrix of the Sierra 2 Line. q Fig. 28 is a partial enlarged view showing an example of a third pattern obtained by the error diffusion method based on the matrix of the Sierra Filter Lite. Fig. 29 is a partially enlarged view showing an example of a third pattern obtained by using the error diffusion method of the Burks matrix. Fig. 30 is a partial enlarged view showing an example of a third pattern obtained by using the error diffusion method of the matrix of Stevenson & Archeg. Fig. 31 is a partially enlarged view showing a second pattern produced by using the third pattern 89 321900 201042295 shown in Figs. 23 to 30. Figure 32 is a comparison of the spatial frequency distribution of the third pattern binarized by the error diffusion method of the various matrices shown in Figs. 23 to 30, and the spatial frequency distribution of the pattern binarized by the threshold method. Figure. Fig. 33 is a diagram comparing the number of generated isolated points when the third pattern is created by the error diffusion method based on the generally known error diffusion matrix, as compared with the case where the threshold value method is used. Figure 34 is an example of an error diffusion matrix showing a diffusion distance of one. Figure 35 is an example of an error diffusion matrix showing a diffusion distance of two. Figure 36 shows an example of an error diffusion matrix with a diffusion distance of three. Figure 37 is a diagram showing an example of an error diffusion matrix with a diffusion distance of four. Figure 38 is an example of an error diffusion matrix with a diffusion distance of 5. Figure 39 is a diagram showing an example of an error diffusion matrix with a diffusion distance of 6. Figure 40 shows an example of an error diffusion matrix with a diffusion distance of 3 + 4. Figure 41 shows an example of an error diffusion matrix with a diffusion distance of 4+5. Figure 42 shows an error diffusion matrix with a diffusion distance of 3 + 4 + 5. 90 321900 201042295 An example of a graph. Fig. 43 is a partially enlarged view showing an example of a third pattern obtained by using the error/diffusion method of the matrix shown in Fig. 34. Fig. 44 is a partially enlarged view showing an example of a third pattern obtained by using the error diffusion method of the matrix shown in Fig. 35. Fig. 45 is a partially enlarged view showing an example of a third pattern obtained by using the error diffusion method of the matrix shown in Fig. 36. Fig. 46 is a partially enlarged view showing an example of a third pattern obtained by using the error ^ diffusion method of the matrix shown in Fig. 37. Fig. 47 is a partially enlarged view showing an example of a third pattern obtained by using the error diffusion method of the matrix shown in Fig. 38. Fig. 48 is a partially enlarged view showing an example of a third pattern obtained by applying the error diffusion method according to the matrix shown in Fig. 39. Fig. 49 is a partially enlarged view showing an example of a third pattern obtained by using the error diffusion method of the matrix shown in Fig. 40. q Fig. 50 is a partial enlarged view showing an example of a third pattern obtained by using the error diffusion method of the matrix shown in Fig. 41. Fig. 51 is a partially enlarged view showing an example of a third pattern obtained by applying the error diffusion method according to the matrix shown in Fig. 42. Fig. 52 is a view showing the number of generated isolated points when the third pattern is created by the error diffusion method based on the error diffusion matrix shown in Figs. 34 to 42 and compared with the case where the threshold value method is used. Figure 53 is a diagram showing the frequency distribution and the threshold value method of the third pattern of the third pattern of the 43th to 51st graphs binarized by the error diffusion method of the error diffusion matrix shown in Figs. 34 to 42. A comparison of the spatial frequency distributions of the binarized patterns. Fig. 54 (4) to (c) are diagrams showing an example of a method of processing an isolated point in accordance with the Monte Carlo method. Fig. 55 (4) to (1) show a graph showing changes in the fourth pattern caused by the number of applications of the Monte Carlo method. Figure 56 shows the relationship between the number of applications of Monte Carlo and the number of isolated points. Fig. 57 (4) to (e) are diagrams schematically showing a preferred example of the first half of the mold manufacturing method of the present invention. Fig. 58 (4) to (c) are diagrams schematically showing a preferred example of the latter half of the mold manufacturing method of the present invention. The first 赖 赖 丨 丨 丨 丨 丨 丨 丨 丨 丨 丨 丨 丨 丨 丨 丨 丨. Fig. 60 (a) and (8) are diagrams schematically showing the state in which the uneven surface formed by the i-th step is subjected to the second rhyme step. Fig. 61 is an enlarged view showing a unit portion to be used in the embodiment i. The image of the unit pattern of the first to third embodiments used in the first embodiment is shown in a partial enlargement. Fig. 63 is an enlarged view showing a portion to be used in the second embodiment. The solid state is shown in the figure. The figure used in the third embodiment is enlarged. The singularity of the singularity of the singularity of the first embodiment is shown in Fig. 65. • Fig. 66 shows an enlarged view of the unit pattern to be used in Comparative Example 2. Fig. 67 is a partially enlarged view showing the unit pattern to be used in the fourth embodiment. Fig. 68 is a view showing the spatial frequency distribution of the unit pattern used in the first to third embodiments. ® Figure 69 shows the spatial frequency distribution of the unit pattern used in Comparative Examples 1 to 2. Fig. 70 is a view showing the spatial frequency distribution of the unit pattern used in the fourth embodiment. Fig. 71 is a partially enlarged view showing the unit pattern to be used in the fifth embodiment. Fig. 72 is a view showing the spatial frequency distribution of the unit pattern used in the fifth embodiment. Fig. 73 (a) and (b) are diagrams schematically showing the evaluation method of flicker. Fig. 74 is a partially enlarged view showing the unit pattern to be used in the sixth embodiment. Fig. 75 is a view showing an enlarged view of the unit pattern to be used in the seventh embodiment. Fig. 76 is a graph comparing the spatial frequency distribution of the pattern shown in Fig. 74 with the spatial frequency distribution of the pattern shown in Fig. 75. Fig. 77 is an enlarged view showing the unit pattern to be used in the embodiment 8 in the section 93 321900 201042295. Fig. 78 is an enlarged view showing a unit portion to be used in the ninth embodiment. The spatial frequency distribution and the spatial frequency distribution of the pattern shown in Fig. 78 of the pattern shown in Fig. 77 are compared with Fig. 79. The first pattern used in the production of the pattern 1 - Fig. 81 is a partial enlarged view of the pattern 1. Fig. 82 is a view showing a partial enlargement of the pattern 2. Fig. 83 is a view showing a partial enlargement of the pattern 3. Fig. 84 is a partially enlarged view showing the first use of the pattern 4. The Fig. 85 of the national system shows a partial enlargement of the pattern 4. Fig. 86 is a view showing a partial enlargement of the pattern 5. Fig. 87 is a view showing a partial enlargement of the pattern 6. Fig. 88 is a partial enlarged view showing the first use of the pattern 7.茱°予

The first pattern D is shown in Fig. 89 as a partial enlargement of the pattern 7. Fig. 90 is a view showing a partial enlargement of the pattern 8. Fig. 91 is a view showing a partial enlargement of the pattern 9. Fig. 92 is a view showing a partial enlargement of the production of the pattern 10. Fig. 93 is a view showing a partial enlargement of the pattern 10. Fig. 94 is a view showing a partial enlargement of the pattern 11. 321900 94 201042295 Figure 95 shows a partial enlargement of the pattern 12. Fig. 96 is a view showing a partial enlargement of the i-th pattern E used for the production of the pattern 13. Fig. 97 is a view showing a partial enlargement of the pattern 丨3. Fig. 98 is a view showing a partial enlargement of the pattern 14. Fig. 99 is a view showing a partial enlargement of the pattern 15. 〇 〇 Fig. 100 shows the spatial frequency distribution of the first patterns A to E. Fig. 101 is a view showing the spatial frequency distribution of the patterns 1 to 3. Figure 102 shows a plot of the spatial frequency distribution of patterns 4 through 6. Fig. 103 is a view showing the spatial frequency distribution of the patterns 7 to 9. Fig. 104 is a diagram showing the spatial frequency distribution of the patterns 1 〇 to 12. Figure 105 is a diagram showing the spatial frequency distribution of the pattern 13 Magic 5. Fig. 106 is a diagram showing the degree of reduction in frequency components due to the patterning method. Figure 1〇7 shows the relationship between the method of creating a pattern and the number of isolated points. Figure 108 shows a partial enlargement of the non-frequency gentleman's case. Figure 1 of Figure 10 shows the distribution of the 丨(10) _ and the borrowing method. Figure 110 shows that the S 1 pattern shown in Figure # will be large. Figure. The pattern of the value is placed 321900 95 201042295 The 111th picture shows the picture of the high-pass filter, the error expansion I, and the pattern will be enlarged. Disc 豕 卡罗 卡罗 第 图 图 图 图 图 图 图 图 图 图 图 图 图 图 图 图 图 图 图 图 图 图 图 图 图 图 图 图 图 图 图 图 图 图 图 图 图No, the isolated point of the pattern is a distribution map. Figure 113 compares the spatial frequency of the pattern shown in (10) to (1). [Key element symbol description] 1 point 8 Surface 10, 11 area 13 Unmasked part 15 First surface uneven shape 17 Key surface 60 Unit 62 Opening portion 65 Light box 67 Glass plate mold base material 9 Photosensitive resin film 12 Mask 14 Dotted line 16 Mine ^ Ming &quot; Layer 18 Second surface uneven shape 61 Chrome light-shielding pattern 63 Photomask 66 Light source 70 Anti-glare film 32 !9〇〇96

Claims (1)

201042295 VII 1. Application scope of Shenyi: - Anti-glare treatment method for transparent substrate, which has: a plurality of points arranged in the ground, or a brightness distribution = 囷 case, the application of the wave from the first case The spatial frequency included is: eight, a step of forming a second pattern by dividing the blade with a low spatial frequency whose spatial frequency is not up to a specific value; and Ο Ο processing the uneven shape on the transparent substrate according to the second pattern 2. For the anti-glare treatment method according to item i of the patent application, = is the solution from the space contained in the first pattern = (4) the low-rate _ rate component of which the rate does not reach a specific value, and the second item In the anti-glare processing method, the above-mentioned killing system is only from the spatial frequency component included in the first pattern; the frequency of::: is not reached. A 'low spatial frequency into the heart 4' method, in which the space of the first pattern of the first money is reduced: the low-frequency component of the frequency between f and the specific value is not: Exceeding the extraordinarily high spatial frequency components, and thus extracting the band-passing data of the spatial frequency components of the ruin of the temple. ·=Please take the anti-glare treatment method of item 4 of the patent scope::::= The lower limit of the inter-work frequency β of the above-mentioned specific range of two extracted is 〇〇ΐβ or more, and the upper limit is 321900 97 201042295 The value T is l /(Dx2)/zm 1 or less [D(#m) is the resolution of the processing apparatus used when processing the uneven shape on the transparent substrate]. 6. The anti-glare treatment method according to Item 5 of the patent application, wherein the upper limit value T and the lower limit value B of the spatial frequency satisfy the following formula (1): 〇.20 &lt; 2x(TB)/( T + B) &lt; 0.80 (!). 7. The anti-glare processing method according to claim 2, wherein the method further comprises: applying a dithering method to the second pattern to produce a third pattern converted into discretized information, And performing the step of processing the uneven shape on the transparent substrate according to the third pattern. The anti-glare treatment method of claim 7, wherein the aforementioned dithering method is an error diffusion method. The anti-glare processing method according to the eighth aspect of the patent application of the present invention, wherein the third pattern is produced by using an error diffusion method in which a conversion error is spread on a three-image thin surface and a six-image phase. The anti-glare treatment method of claim 7, wherein the third pattern is converted into a pattern of information that is discretized in two stages. ^ The anti-glare processing method of claim 10, wherein the method further comprises: moving the isolated black or white pixel by using the third pattern by the geo-calculus method converted into the two-stage discretized information In the step of 4 patterns, through the upward direction: step: the fourth pattern is processed to form the concave and convex shape on the aforementioned concave and convex portion 12. The anti-glare treatment method as in the application of the first item, t, 321900 98 201042295 The transparent substrate? The old (four)...-cup step comprises: forming a mold having a concave-convex surface according to the first pattern _. ^ VIII ugly transferring the concave and convex surface of the mold to a transparent substrate Step by step. 13. The anti-glare treatment method according to claim 7, wherein the step of processing the uneven shape on the ruthenium substrate comprises: forming a mold having a concave-convex surface according to the third pattern, and the mold is The step of transferring the uneven surface onto the transparent substrate. 14. The anti-glare treatment method according to claim 5, wherein the step of processing the uneven shape on the transparent substrate comprises: forming a mold having a concave-convex surface according to the fourth pattern, that is, the unevenness of the mold The step of transferring the surface onto the aforementioned transparent substrate. 15. The anti-glare treatment method according to claim 7, wherein the step of processing the uneven shape on the transparent substrate is a processing device that performs processing based on the discretized information of the third pattern. And proceed. In the anti-glare treatment method of claim 11, wherein the step of processing the uneven shape on the transparent substrate is performed by using the discretized information according to the fourth pattern. The device is processed. A method for producing an anti-glare film includes: applying a filter to a first pattern of irregularly arranged or having a luminance distribution, at least removing or reducing a space from a spatial frequency component included in the first pattern a step of producing a second pattern with a frequency of a low spatial frequency component having a specific value; and 321900 99 201042295 a step of processing the uneven shape on the transparent substrate according to the second pattern. The method for producing a film according to the fourth aspect of the invention, wherein the chopper device removes or reduces a low spatial frequency component having a spatial frequency that does not reach a specific value from a spatial frequency component included in the first pattern. High pass filter. 19.1. The method of claim 18, wherein the chopper is to remove or reduce a low spatial frequency of a spatial frequency less than G.G1 from the spatial frequency contained in the first pattern: High-pass filter. The method for producing an anti-glare film according to the invention of claim 17, wherein the filter is a spatial frequency component included in the first pattern; and a low spatial frequency component having a frequency of eight or less is not a specific value: The dream space frequency exceeds a high spatial frequency component of a specific value, and the second b extracts a band filter of a specific fan_empty_rate component. The manufacturing method of the anti-glare film of claim 20, wherein the spatial frequency lower limit value β of the specific range of the above-mentioned specific range extracted by the band pass chopper is 'the upper limit value τ is i /(Dx2)〆The following [D force used when processing concave and white on a transparent substrate]. Analysis of the apparatus 22. The manufacture of the anti-glare film of claim 21;;) the upper limit τ and the lower limit B of the spatial frequency are satisfied;: 321900 100 201042295 0. 20&lt;2χ( Τ-Β)/(Τ + Β)&lt;0. 80 (1). 23. The method for producing an anti-glare film according to claim 17, wherein the method further comprises: applying a dithering method to the second pattern to produce a third pattern converted into discretized information; And performing the step of processing the uneven shape on the transparent substrate according to the third pattern. 24. The method for producing an anti-glare film according to claim 23, wherein the dithering method is an error diffusion method. 25. The method for producing an anti-glare film according to claim 24, wherein the third pattern is produced by applying an error diffusion method in which a conversion error is spread in a range of 3 pixels or more and 6 pixels or less. 26. The method of producing an anti-glare film according to claim 23, wherein the third pattern is converted into a pattern of information that has been discretized in two stages. 27. The method for producing an anti-glare film according to claim 26, wherein the method further comprises: for the third pattern converted into two-stage discretized information, the isolated black is made by Monte Carlo method Or a step of moving the white pixel to form the fourth pattern, and performing the step of processing the uneven shape on the transparent substrate according to the fourth pattern. 28. The method for producing an anti-glare film according to claim 17, wherein the step of processing the uneven shape on the transparent substrate comprises: forming a mold having a concave-convex surface according to the second pattern, and The step of transferring the uneven surface of the mold onto the transparent substrate. The method for producing an anti-glare film according to claim 23, wherein the step of processing the uneven shape on the transparent substrate comprises: forming a mold having a concave-convex surface according to the third pattern; And the step of transferring the uneven surface of the mold onto the transparent substrate. The method for producing an anti-glare film according to claim 27, wherein the step of processing the uneven shape on the transparent substrate comprises: forming a mold having a concave-convex surface according to the fourth pattern, and The step of transferring the uneven surface of the mold onto the transparent substrate. The method for producing an anti-glare film according to claim 23, wherein the step of processing the uneven shape on the transparent substrate is performed by using the discretized information according to the third pattern. The device is processed. The method for producing an anti-glare film according to claim 27, wherein the step of processing the uneven shape on the transparent substrate is performed by using the discretized information according to the fourth pattern. The device is processed. 33. A method of manufacturing a mold for manufacturing a mold described in claim 12, 13, 14, 28, 29 or 30, the method comprising: performing on a surface of a substrate for a mold a first plating step of copper plating or nickel plating; a polishing step of grinding a surface coated with a plating step by the first plating step; and a photosensitive resin film 102 321900 201042295 formed by forming a photosensitive resin film on the polished surface a step of exposing the second pattern, the third pattern, or the fourth pattern to a photosensitive resin film; and exposing the photosensitive resin to the second pattern, the third pattern, or the fourth pattern a developing step of developing the film; a first etching step of forming an unevenness on the polished plated surface by using a developed photosensitive resin film as a mask; and a photosensitive resin film peeling step of peeling off the photosensitive resin film; And a second plating step of chrome plating on the formed uneven surface. The method for producing a mold according to claim 33, wherein the photosensitive smear film peeling step and the second plating step include etching treatment to form irregularities of the uneven surface The shape is gentle and the second etching step. The method of producing a mold according to claim 33, wherein the embossed surface formed by chrome plating formed in the second plating step is a mold uneven surface on which the Q is transferred onto the transparent substrate. The method of manufacturing a mold according to claim 33, wherein the chrome plating layer formed by the chrome plating in the second plating step has a thickness of 1 to 10 // m. An anti-glare treatment method for an image display device, which is characterized in that the anti-glare treatment is performed on the surface of a transparent substrate provided in the image display device by the anti-glare treatment method described in the first paragraph of the patent application. 38. An image display device comprising an anti-glare film obtained by the method for producing an anti-glare film according to claim 17 of the patent application. 103 321900