TW528983B - Dithering process device and method in image process and computer graphic system - Google Patents

Abstract

The present invention discloses a dithering method for converting N-bit digital image data into M-bit digital image data, where N-M=K, K > 0. The method can be applied to a dithering method in the image process system and computer graphic system. For example, the present method can convert the N-bit red color value of each pixel into N-bit red color value only when there is very small or no color deviation. The present method is based on the difference between the image data of the most significant n bit and the least significant n bit of the pixel to provide a predicted standard, where n=k in the best condition. In addition, the present invention also discloses a dithering device for converting N-bit digital image data into M-bit digital image data, where N-M=K, K > 0. A subtracting operation device included in the device is provided to obtain a difference by subtracting the least significant n bit from the most significant n bit of the pixel of the N-bit image data.

Description

528983

[Scope of application of the invention] The present invention is about a kind of N-bit image data of a digital image, and a flutter color mixing to M-bit 70 image data) method, where NM = K, and K > 0, special This method is applied to the dithering process of computer graphics or image processing to solve the problem of image deviation. [Related Technology and Background of the Invention] In computer graphics or image processing, high-precision images are often converted to low-level images. Generally speaking, the processing benefits are capable of processing high-level image data. , But the next storage device or display device must be compatible. However, sometimes it must be limited to low-level standards. This is why most image processors are designed to support different color-level display standards. For reasons such as 24-bit (true) color true mode j (true c0l0r f0rmat) and 16-bit (Mt) high-level color invariants (In gh color f0rmat), therefore, 24 The conversion of bit materials into 16-bit data to facilitate storage or display is often used. In the 24-bit true color mode, 8-bit is red, 8-bit is green, 8-bit is blue, and 16-bit high-level color mode uses 5-bit, 6-bit and 5-bit to represent red, green, and blue 'for now' When you want to convert a 24-bit true color mode pixel into a 16-bit high-level color mode, you need to apply The conversion algorithm of 8-bit to 5-bit and 8-bit to 6-bit shown in the first figure.

528983 V. Description of the invention (2) It is known that the existing truncation method (truncat i on me thod) and the rounding method (rounding method) are applied to the above-mentioned conversion algorithm, but the simple truncation method and rounding method are used because the final image will be Mach banding effect is generated but is not very reused. The above-mentioned Mach banding effect will cause an intensity gap at the boundary, that is, there is a discontinuity in the magnitude and change of the gradation value. This situation can be seen visually; in order to avoid the Mach band effect, it has been known that f: a kind of dithering color mixing method to improve the above situation, the dithering color mixing method, the color scale value obtained by the truncation method ( c〇1〇r value) combined with a dithering color mixing value, where the dithering color mixing value changes with the position of the pixel = the same, the three figures shown in the second A to 20 Since the bit position is to reduce dither matrix and dither table when 2-bit, 3-bit and 4-bit are reduced, its% and j represent the coordinates of the X direction and Y direction of the pixel, respectively. The remaining value after m is expressed as 1 = (乂 m〇dm), bu (ym〇d ...) According to the use of the dithering color mixing table, add some random values (rand〇m = this) _ after adding to the pixel These flutter color mixing effects will clip the outline: the random numbers in the U color mixing table are calculated by different algorithms. The right and right pictures are added to this flutter color mixing effect, which is caused by the color level value. Boundary will be softened. /, Ba 1 white processing is a series of applications. First, silly I Ϊ refer to / 3 pictures, which are the functions of the conventional tremor blending method. The block diagram transforms t from the true color mode into a high-resolution color mode: the seat of the pixel; the f value (that is, and 乂) is entered into the function block

528983 V. Description of the invention (3) Find a comparison value T in the dithering color mixing table of D3 described in the second figure B. In function block 32, compare the comparison value with the pixel's The value of the lowest 3 bits (R [2 ·· 0]) of the 8-bit color gradation value (R [7: 〇]) is compared to determine a carry value. If the value is the lowest 3 bits (R [2: 0]) is greater than the comparison value, its carry value is set to 1, in function block 33, the highest 5 bits of the 8-bit color scale value (R [7: 0]) in the pixel The value of (R [7: 3]) is increased by 1, and expressed as a temporary value (Tmp [5: 0]); Conversely, if the value of the lowest 3 bits (R [2 ·· 0]) is less than or Equal to the comparison value, the value of the last 5 bits (R [7: 3]) of the 8-bit color scale value (R [7: 〇]) in the pixel is added to 0, that is, the pixel The value of the highest 5 bits (r [7: 3]) of the 8-bit color gradation value (R [7: 0]) is directly expressed as a temporary value (Tmp [5: 0]); in function block 34, Furthermore, the temporary value is converted into a 5-bit color finally required by the pixel by using a clamp 〇perati〇n method. In the above overflow processing method, if the temporary value is greater than 31, the color level value (R '[4: 0]) after the dithering is set to 3 1, otherwise the color level after the dithering is mixed The value (R '[4: 〇]) is equal to the temporary value (Tmp [5: 〇]). The above is to explain that when the temporary value adds a carry value and an overflow occurs, make 52 = value (Tmp [4 : 〇]) becomes a 6-bit temporary value (;; processing status. "" Rishou please refer to the fourth figure, $ diagram is similar to the third. = Law "processing-8-bit pixels are converted into 6-bit The color block diagram of the conversion algorithm is used to convert green from the true color mode to j = square mode. Similarly, first input an image +, and the coordinate values (ie, X and y) of the elbow pixels to resolve the color

528983 V. Description of the invention (4) A comparison value (threshold vaiue) τ is found in the free-vibration color mixing table of j) 2 in block 41 of Liyue b. 41, in function block 42, After comparing the comparison value with the value of the lowest 2 bits (G [1: 〇]) of the 8-bit color scale value (G [7 ·· 〇]) of the pixel, a round value (v ^ lue) 'If the value of the lowest 2 bits (G [1: 〇]) is greater than the comparison value, the carry value is set to 1. In function block 43, the 8-bit color scale value in the pixel (G [7 ·· 0]) The value of the highest 6 bits (G [7 ·· 2]) is increased by 1 and then expressed as a temporary value (Tmp [6: 〇]); on the contrary, if it is the lowest 2 The value of bit tl (G [1: 〇]) is less than or equal to the comparison value, and the value of the highest 6 bits (G [7: 2]) of the 8-bit chroma value (R [7: 0]) in the pixel is The value is added, that is, the value of the most significant bit (G [7.2]) of the 8-bit color scale value (G [7: 〇]) in the pixel is directly expressed as a temporary value (Tmp [6: 〇] 'In function block 44, the temporary value is further converted into the final required 6 of the pixel by overflow processing. The value of the color level of the element; the overflow of the above description: When the temporary value is greater than 63, the color of the dithered color \]) is not set to 63, otherwise the color level of the dithered color is temporarily added to the carry value and overflow occurs. After the bit, make-"Li:"; =. [5: °]) becomes a temporary value of -7 bits (T ·.]). Dynamic-: The first or the fourth image conversion algorithm is called the inverse dither. Γ is often used for high-level color The mode is converted to "beige-like". The general inverse chatter is mixed with "time" as described in block 51_.

Page 8 528983 V. Description of the invention (5) Bit S In the fifth A chart, the 5-bit and color gradation values (R '[4: 0]) in the original high color gradation color mode are directly used as the true color mode. The highest 5 bits: color & step value U '' [7: 3]), and then the lowest 3 bit step value of the 5 bit color step value, (R '[4: 2]) as the true color mode The lowest 3-bit color-gradation value-(R '' [2: 〇]). The combination of the above two values returns to the 8-bit color-gradation value (R [7: 0]) of the true color mode; similarly, in In the fifth β image, the 6-bit color gradation value (G, [5: 〇]) in the original high-level color mode is directly used as the highest 6-bit color gradation value (Q, , "7 · 2 ί> The highest 2-bit color-gradation value (U) 2U-color mode and the 2-bit color-gradation value (G ,, [丨: 〇]), The combination of the above two values returns an 8-bit color gradation value (G ,, [7: 〇]) of the true consistent color mode. The above-mentioned dithering color mixing method has directly obtained the gradation conversion by using the truncation method. Better visual quality, however, in computer graphics applications, The image after the dithering process often needs to be inversely dithered, and then it is dithered again. The conventional method of inverse dithering is to add a bit of high-order color levels to the dithered color. In the color scale value, 'to convert the pixels from low color scale to high color scale, but it is very bad that some of the points in the final image will appear in the process of repeating the process of dithering and inverse dithering repeatedly like this. There is a phenomenon of too red, too green, or even too blue. This is what causes the so-called color deviation. The following explains how the conventional dithering method produces color error problems. · Suppose there is a pixel at a certain point. Dithered color mixing and inverse dithering color mixing several times, the number of color level bits of this pixel is set to N, the bits that have been dithered color mixing

528983 V. Description of the invention (6) According to the position of the pixel 'through a dithering color mixing table τ, the gradation value of the second / first step is not B, after k times of dithering color mixing ^ ^ ^, θ ρ & value Set it to L, this pixel undergoes k dithering process A _ ". The value of M is the lower η bit value, so we can know that p-P f is over the conventional method of inverse dithering, and we get +, which is Uk_ ; In the first dithering color mixing program, if P. is greater than T, then ν1 v ° = and Il = M8 + Hl, that means Pl = Hl; if Pl is greater than τ ', then that :: 1 1. This It ’s easy to prove that if H1 is greater than or equal to H1, if Hl is greater than T, Hk will be greater than pk, and pk will be greater than T. ^ 目 Θ ^ Λ2k will also increase. Finally, we will find that These points start to appear inconsistent colors, that is, color errors. How to improve this — the lack of means is the main purpose of developing this case. [Summary of the Invention] Therefore, the main purpose of the present invention is to provide a method and device, Regarding the conversion of N-bit image data of digital images to M-bit image data /, where N-M = κ, And κ> 0, to solve the problem of image deviation. The first implementation concept provided by the present invention is a dithering method for converting ^ bit image data of a digital image into M bit image data, where N- Μ = κ, and κ >〇; The method includes the following steps: Execute the first operation to convert the bit of the digital image data to the bit 528983

A difference between the most significant bit and the lowest n bits in the N-bit image data is obtained; then a second operation is performed to convert the pixel coordinate value to a comparison value; after comparing the difference value and the comparison value When the absolute value of the difference is not greater than the comparison value] set the highest M bit of the N-bit image data of the pixel to the μ-bit image data of the pixel; when the absolute value of the difference is greater than the comparison value and the difference When the value is a positive number, the highest% bit of the N-bit image data of the pixel is added to the first bit value as the M-bit image data of the pixel; finally, when the absolute value of the difference is greater than the comparison value-and the difference is When the value is a negative number, the highest M bit of the N-bit image data of the pixel is reduced by a second value, and the M bit image data of the pixel is used. According to the above conception, in the method of flutter color mixing, i η N / ·, and Wei η ^ κ. According to the above-mentioned concept, the first value and the second value in the method of dithering and color mixing are equal and both are equal to 1. / According to the above concept, the method of dithering color mixing further includes a step: comparing the highest 1 bit of the N-bit image data with the sum of the first value (2M -1). If the above sum is greater than (2LD, Then (2M-n); f-bit M-bit image data. According to the above conception, the method of dithering and color mixing further includes a step... The highest M-bit of the N-bit image data and the second value. The difference is compared with 0. If the above difference is less than 0, then 0 is regarded as the M-bit image data of the pixel. According to the above conception, the first tube described in the method of mixing colors includes a subtraction operation; and The second operation includes a dithering operation; (dithering operation) °

Page 11 528983 V. Description of the invention (8) According to the above idea, the image data of the pixel described in the method of dithering color mixing is converted according to the red, green and blue color level values of the pixel; further, the above The steps can be applied to other pixels of the digital image, and the N-bit image data of the digital image is converted into μ-bit image data. The second implementation concept provided by the present invention is a device for a dithering process, which is used to convert N-bit image data of a digital scene > image to μ-bit image data, where N-M = Κ, and κ> 0; the device includes at least a subtraction device for subtracting the highest n bits of the N-bit image data of the pixel from its lowest n bits to obtain a difference; a comparison value generating device , Used to obtain a comparison value after the pixel coordinate f;-a comparison operation device is electrically connected to; a method for generating different values and a comparison value generation device for comparing the difference and the comparison value. When it is not greater than the comparison value, the comparison operation device outputs the 丄 ith system, the absolute value of the difference value is greater than the comparison value, and the difference value is a comparison operation device, and the second control signal is output; when the difference is two, the difference value and the difference value When it is a negative number, the comparison operation device wheel: the upper part: the calculation device is electrically connected to the comparison operation device, and the calculation device is outputted as the M-bit image data of the pixel; the calculation bit & control signal converts the M bit of the pixel The image data is based on the second above. M-bit image data output as pixels ;: M-bit image data of pixels was installed with 70 plus the second-most control signal; 'According to the above--the value of the M-bit image data wheel of pixels Out. Ask *

528983 V. Description of the invention (9) According to the above idea, it should be applied to the value of the more plain M-bit image data in the device which further includes an overflow processing device, Ray = two one sequence, and (with-different device, use If the value of the image data of the pixel is greater than (2m ~ 丨), 3% is enabled, and the above-mentioned M adjustment operation is performed; if the above-mentioned bit is changed, the processing device executes the first processing device and performs the second adjustment. If the value of the material is less than 0, the value of the μ bit image data # of the pixel at the overflow is replaced by the value of the image data of the first bit becomes the M bit image data ::::; Based on the above-mentioned idea, the M-bit image data of the pixel. According to the above idea, the comparison value generating device for performing dithering includes-dithering color mixing i = her = r), and is used for generating the comparison of the coordinate values of corresponding pixels. Value, up, and down, to achieve the comparison value according to the dithering color mixing table described above ... To apply the practice [explanation of the embodiment] Please refer to the sixth diagram, this diagram shows the application of the present invention to a two = loose conversion into High-level color mode. The device includes a subtraction nosepiece set 11, a comparison value generating device 丨 2, a comparison operation device 丨 3, a setting 14, and an overflow processing device 15, #. Among them, the comparison value generating device 12: '/ one can reduce 3 bits Yuan's dithering color mixing table (D3) 〇 ^ 'First, input the coordinate value of one pixel (that is, \ and 7) into the comparison value Yusheng device 12, as shown in the second dithering D3's dithering color mixing table. Find the comparison value (threshold value) T, meanwhile, the highest 3 bits (R [7: 5]) and lowest of the color scale value of the pixel on page 13 ^ 28983, invention description (10) " '' After 3 bits (R [2: 〇]) are passed through f subtraction operation device 1 丨, a difference value (D [3: 〇]) is obtained, and the difference value and comparison value T are compared with the comparison operation device 丨 3. , Output an operation value a (〇P [1: 0]); if the absolute value of the difference (D [3: 0]) is less than or equal to the comparison value T, the operation value (OP [1: 〇] ) Will be set to 0, which means that no change is necessary; on the contrary, if the absolute value of the difference (D [3: 〇]) is greater than the comparison value T, the calculated value (0P [丨: 〇]) will be Will be based on the sign of the difference (D [3 ·· 〇]) Depending on the number, when the sign of the difference (D [3: 〇]) is positive, the calculated value (〇P [1: 〇]) is set to 1; on the contrary, when the difference (D [3: When the sign of 0]) is negative, the operation value (〇P [1: 0]) is set to 2; when the above operation value (〇p [丨 ·· 〇]) determines the 'a temporary value ( Tmp [5: 〇]) can be calculated as shown in Equation 14 below: ^

Tmp [5: 0] R [7: 3] if OP = 〇r [7: 3] + 1 if 〇P zz 1, R [7: 3] -1 if OP = 2 τγ value (Tmp [ 5: 〇]) and then adjust it to a 5-bit color gradation value through the overflow processing device I. The adjustment equation is as follows: R, [4: 0] 31 if Tmp [5: 〇] > 31 · 0 if Tmp [5: 〇] < 〇, Tmp [5: 〇] After others, 'If there is a need to convert the high-level color mode to the true color mode, you can consider the inverse flutter described in Figure 5A Color mixing method

Page 14 528983

5. Description of the invention (11) Like reply. The seventh figure shows the preferred embodiment of the present invention applied to a pixel that converts 8-bit to 6-bit. The dithering color mixing method is similar to that described in the sixth figure, and is used to convert the green color level from the real color mode to high. Levels color mode. First enter the coordinate value of a pixel (that is, X and y) into the function block 71, and determine a threshold value T ′ from the dithering color mixing table of D2 as described in the second A figure. At the same time, 'Subtract the highest 2 bits (R [7: 6]) and the lowest 2 bits (R [1: 0]) of the pixel's color scale value from the function block ^' to obtain a difference (D [ 2 · 0]) 'The difference value and the comparison value τ are input into the function block 7 3 After comparison, an operation value (〇p [丨: 〇]) is output; if the difference value (D [2 ·· 0]) When the absolute value is less than or equal to the comparison value _, the calculated value (0P [1: 0]) will be set to 0, which means that no change is necessary; on the contrary, if the difference (D [2: 0] When the absolute value of) is greater than the comparison value T, the calculated value (OP [1: 〇]) will be determined according to the sign of the difference (d [2: 〇]). 'When the difference (D [2: 0]) When the sign is positive (p0siHve), the value of the fortune management (OP [1: 0]) is set to 1; on the contrary, when the sign of the difference (j) [2: 〇]) is negative, the operation value ( 〇P [1: 〇]) is set to 2; after the above calculation value (〇P [1: 〇]) is determined, a temporary value (Tmp [6: 〇] ) Can be different through function block 7 4. The calculation formula of this process is as follows: | G [7: 2] if OP = 〇

Tmp [6: 0] = jG [7: 2] + l if OP = 1, G [7: 2] —1 if 〇P = 2

528983 V. Description of the invention (12) — The value of the flutter color mixture is obtained by processing the temporary value (Tmp [6: 〇1, ^ Block 75 overflow), and its equation is as follows: : 63 if Tmp [6: 〇] > 63

Gr [5: 〇]-0 if Tmp [6: 〇] < 0

Tmp [6: 〇] Others According to the above embodiments, it can be understood that the method of blessing color according to the present invention is based on Wh + ι, or vVy, which is based on the criteria of the color movement, and the inverse flutter color mixing method is applied The conventional conversion invention of the dithering color mixing method is to first calculate and compare with the "pan", "1", 1 ^ comparison · τ and determine whether the dithering coloration value is "1" or "maintaining change". "," Minus $ In each conversion criterion, it can be clearly seen that p is derived from the inverse dithering color mixing method, that is, the difference between the 1st and ^ undertremor 'and "after the color system" Will become 0 'in other words', that is, p "Hk = 〇, without any change, Ik + 1 will still be equal to Ik, and the flutter color mixing error value table (UUher error) will not be expanded, of course, to avoid. So The main purpose of effectively achieving the case can be described above. It is only for comparison of the present invention. Each series is used to explain rather than to limit the book. .Everything according to the present invention. The temple culture and decoration should belong to the scope of the present invention. 528983 Simple illustration of the diagram [Simplified illustration of the diagram] The first picture is shown in the conventional computer drawing. Digital image of Hitachi wide-receive true color mode = block diagram of the color program. "The image is converted to the color mode of color scale. The third and second picture A shows three kinds of dithered color mixing array or dithered color mixing: Norm: applied to remove 2 bits (D2) and remove 3 4 (D3) and a conversion method for removing 4 bits (D4). = In the known dithering color mixing program, the process of processing one pixel from 8-bit conversion of the fourth image, and the method of changing the method ® is used to convert the red ... shell shirt mode into a high-resolution color mode: the known dithering color mixing In the program, a function symbol that handles a pixel y into a 6-bit conversion algorithm Xia Xia # ^ # ^ ^ ^ A block diagram of the power moon, which is used to convert the green Λ β form into a high-resolution color mode. The fifth image A and the fifth β image are respectively a m _ ^ & ^ are described in the second image and the fourth image ^ a movement 'the inverse dithering color mixing program of this color program. The sixth figure is a method and device for processing a similar image in the present invention. The method and device for converting ί 5 into a 5-bit conversion is shown in the seventh picture. Α Taxi / Fruit conversion to high-resolution color mode. This is the method of the present invention. Method & π conversion into 6-bit conversion straight real i, which is a preferred embodiment, which is used to convert green color to a high-resolution color mode. [Illustration of Symbols] 11 Subtraction Device Page 17 528983

Page 18

Claims (1)

528983 6. Scope of patent application [Scope of patent application] Kind of N-bit image data of one digital image is subject to $ M 彳 -D / multi-data dithering (di th · ,, " transfer to M position A shirt-like method of ancient soil 6 into (g) 'where N-M = κ, and κ > 0, the method includes the following steps: the digital image carefully ## i i 用, using Γ 之 N-th pixel A difference between the most significant bit and the lowest n bit το in the image data of the first L-bit-transmitted bit image data; take a calculation to convert the coordinate value of the pixel into a -comparison value ; Compare the difference between the driver and the comparison value; when the absolute value of the difference is not greater than the comparison value, the highest M bit of the N bit ^ data is set to the pixel ^ bit image breeding; when the When the absolute value of the difference is greater than the comparison value and the difference is a positive number: add a first value to the highest M bit of the N-bit image data of the pixel as the M-bit image of the pixel Data •, and when the absolute value of the difference is greater than the comparison value and the difference is a negative number, 'the N bit of the pixel The highest M bit of the image data is subtracted by a first value from the M bit image data of the pixel. 2 · The method of flutter color mixing as described in item 1 of the patent application range, i ^ "(N / 2). — 3. The method of flutter color mixing as described in item 1 of the scope of the patent application, wherein η 二 Κ. ', Page 19 528983 VI. Patent application scope 4. The method of flutter color mixing as described in item 1 of the patent application scope, wherein the first value mentioned is 1; the second value mentioned is also 1. 5. The method of flutter color mixing according to item 1 of the scope of patent application, wherein it further comprises a step: the sum of the highest M bit of the N-bit image data and the first value and a value (2M- 1) For comparison, if the above-mentioned sum is greater than the value (2M-1), the value (2M-1) is regarded as the M-bit image data of a pixel. 6. The method of flutter color mixing according to item 1 of the scope of patent application, further comprising a step of comparing the difference between the highest M bit of the N-bit image data and the second value with 0, if the above If the difference is less than 0, 0 is regarded as the M-bit image data of the pixel. 7. The method of flutter color mixing according to item 1 of the scope of patent application, wherein the first operation includes a subtraction operation. 8. The method of dithering color mixing as described in item 1 of the scope of patent application, wherein the second operation described by the user includes a dithering color mixing operation. ® 9. The method of dithering color mixing according to item 1 of the scope of patent application, wherein the image data of the pixel can be a combination of the pixel's red, green, and blue color level values. Page 20 528983 VI. Application scope of patents-1 〇 · The method of dithering color mixing as described in item 9 of the scope of patent application, where the steps described can be applied to other pixels of the digital image, and complete the The N-bit image data of the digital image is converted into M-bit image data. The device is used for converting 11-bit N-bit image data to digital image data. , Where NM = K, and KM); The device includes at least: 八, 减, the difference device, which is used to subtract the lowest ^ bit from the southern n bit of the one-bit N-bit image data, Obtain a difference and a comparison value generating device, which is used to obtain a comparison value after calculating the coordinate value of the pixel; &-a comparison operation device, which is electrically connected to the subtraction operation device and generates the comparison value A device for comparing the difference with the comparison value; when the absolute value of the difference is not greater than the comparison value, the comparison operation outputs a first control signal; when the absolute value of the difference is large: ^ Logarithmic and the difference is positive When the number is counted, the comparison operation device has a second control signal; when the absolute value of the difference is greater than the ratio = the difference value is a negative number, the comparison operation device outputs a signal; and the two-control-one calculation device, It is electrically connected to the comparison operation device, and according to the first control signal, the heart element image of the pixel is set to the south M bit, and the M bit image data of the pixel is rotated out. 528983 The computing device is based on the highest M bit 2 of the second image data / alignment element of the pixel ... image data: ς: upper; second:; after the value is regarded as the control signal of the image together, I, out, ° The juice difference device subtracts a second value from the 3 Hai pixel bit shadow field according to the third city J Song 琨 a, and then goes to I to get the 鬲 M bit. Μ-bit image data wheel with value of n Hai pixels 12. The device for performing the procedure as described in item η of the patent application scope, the device further includes a vanadium 仃 a moving the color processing device, which is electrically connected to the The computing device is used to compare the value of the M-bit TG image of the pixel with a value (2M- 丨) or 0. If the value of the above-mentioned M-bit image data is greater than the value (2M-1) , The overflow processing device performs a first adjustment operation; if the value of the M-bit image data is less than 0, the overflow processing device performs a second adjustment operation. ^ 1 3. The device for performing a dithering color mixing program as described in item 丨 2 of the scope of patent application, wherein the first adjustment operation is to replace (2M — 丨) the value of the above M-bit image data It becomes the μ-bit image data of the pixel. ·· 1 · The device for performing a color-mixing program as described in Item 12 of the scope of the patent application, wherein the second adjustment operation is to replace the value of the above-mentioned M-bit image data with the value of 0 M-bit image data of pixels. 528983 15. The device for performing a dithering program as described in item 1 of the scope of patent application, wherein the comparison value generating device includes a dither table, which is used to correspond to pixels. A value of 1 produces this comparison. 1 6. The device for performing a dithering color mixing program as described in item 11 of the scope of patent application, wherein. 1 7 · The device described in item 11 of the scope of patent application is applied to a device that performs a dithered color mixing ^ private sequence, where n = K. 1 8 · The device for applying and performing the tremor heart color program as described in item 11 of the scope of the patent application, wherein the first value described is 1; the second value of the cancer described by the household is also 1. 19. The device for performing the trembling ice y cockroach procedure as described in item 11 of the scope of the patent application, wherein the image data of the pixel is the red, green, and blue color levels of the pixel Values. · Page 23