KR100363965B1 - Method of optimizing color image for 4 gray-level display - Google Patents

Abstract

The present invention relates to a method for displaying a color image on a four-stage color display device such as a 4-gray liquid crystal display device, and a method for outputting an optimized result image by making full use of color information of an original image.

A color image optimization method according to an embodiment of the present invention includes the steps of generating a gray image from an input color image, resetting a divided area based on a color distribution in a gray image, configure the region-specific images with the sub-image _{S 1 (x, y),} S 2 (x, y), S 3 (x, y) step, 3 differentiation of sub-images a _1, a _2, for each of building a , a ₃ values R ₁ and an error filtered by a _{(x, y), R 2} (x, y), R 3 (x, y) a step of generating, R ₂ (x, y) and R ₃ (x , y) converting the image into four gray values, the error filtering step [107] of the generated image R ₁ (x, y) and by merging the translated image of the four gray conversion step last four gray images G (x , y) .

Description

[0001] The present invention relates to a method of optimizing a color image for a gray display device,

The present invention relates to an image optimization method for displaying a color image on a 4-gray display device, and relates to an optimized result image output method that makes full use of color information of an original image.

Modern industries and societies are changing more and more dependent on intangible values of technology and information rather than naturally occurring beings or values. For users who are exposed to a lot of information, acquiring only the necessary information quickly became a means of overcoming competition, and various tools required for such means occurred. The development of computers and networks has played a leading role in the information society, and the desire of more and more users has led to the use of wireless networks to address space and time constraints.

Among the various methods for wireless network, most of users are using easily accessible personal information devices such as mobile phone and PDA. In order to make information more easily and quickly at a glance, It is indispensable. It is the use of a graphic display device having a 4-step color tone developed in one step from a 1-bit monochrome image, in order to simultaneously provide promptness and convenience that satisfies users' desires in utilization of image information.

Currently, the image processing specialist is required for proper valuation of the existing image information for the 4 gray display device, and the time and cost reinvestment is required, which is a constraint factor on the user side of the individual and companies.

4 technology that is used for creating gray image is a method of making drawing by making a user who wants to utilize image information or editing various complicated processing steps one by one by using appropriate image generation tool or software . This requires a great deal of time and money and specialized knowledge of image processing, which hinders the maximization of information utilization and value creation and does not benefit the industry as a whole.

Based on a series of image processing techniques, we devised a new method for effective generation of 4-step tone image and developed a quick and easy software using it. The algorithm and software can be used for effective representation of color images in a graphic display device with a four-step tone.

Accordingly, it is an object of the present invention to provide an image display device and a method for controlling the intensity of the intermediate stage, which are lost when the image is converted from 256 levels of intensity level to 4 levels of gray level, And to provide a method of optimizing a color image for a 4 gray display device that overcomes loss by converting the step change to a density of dots through a value and dithering.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1A is a flow chart illustrating the process of characterizing dithering in accordance with one embodiment of the present invention. FIG.

1B is an algorithm flowchart of a step of generating a gray image.

Figure 1C is an algorithm flow diagram of a partition reset step.

Figures 1D and 1E show an algorithm flow diagram of the overlapping region 3 differentiation step.

1F is an algorithm flow chart of the 4-gray conversion step.

Figure 1G is an algorithm flow diagram of merging sub-images to output the final result.

Figure 2a is a flow diagram of a method for characterizing dithering, according to another embodiment of the present invention.

2B is an algorithm flow diagram of the channel conversion step.

2C is an algorithm flow diagram of the high-pass filtering step.

Figure 2d is an algorithm flow chart of the thresholding step.

3A is a diagram showing a 3x3 neighboring pixel for an arbitrary pixel (x, y). Fig.

3B is a table showing a general 3x3 mask.

FIG. 4 is a table showing a basic highpass spatial filter, which is a general 3 × 3 sharpening filter; FIG.

5 is a table showing masks;

Figure 6 is a graph defining the cut-off frequency of an ideal high-pass filter in a two-dimensional space.

7 is a graph showing characteristics of a Butterworth high-pass filter (BHPF).

8 is an exemplary view showing a resultant image in which a color image is optimized for display on a 4-gray display device according to the present invention.

summary

The color image is the most reliable means of information transmission by showing and conveying the actual image as it is. However, the color image containing the real image has a relatively large data size to be processed from the location of the information to the user acquiring step, and the basic unit of information transmission using the wireless Internet is still the technology and the general utilization method and means It has various prerequisites. It is possible to utilize 256 gray images which can use less data size and low cost display device. However, the high speed processing required on this wireless network and all the information utilization facilities and personal terminals When not appropriate.

Currently, personal wireless terminals and information assistant devices for wireless internet use 1-bit liquid crystal display device which displays two colors of black and white. However, since it is a condition that is too harsh to solve the needs of information users, And a 4-gray display device capable of expressing a four-step hue capable of fast transmission. Therefore, a suitable image information processing tool has become necessary, and an appropriate algorithm and method have been required. Therefore, a program or a method that can be used easily has not been known and it has been newly developed.

Configuration of Embodiment

A color image optimization method for a 4-gray display device can be implemented as " characterization dithering using redundant region 3 differentiation " and " expansion concept of image characterization dithering for a 1-bit monochrome display device ".

(1) Characterization dithering using redundant region 3 differentiation (see Figs. 1A to 2G)

The feature dither method according to this embodiment produces results that are suitable for the conversion of images with a large number of intensity stages of full color images and continuous hues such as photographs.

Referring to Figure 1A,

A step (101) of generating a gray image from the input color image,

(Detect A []) [103], a step of resetting a divided area based on the distribution of colors in a gray image,

A step 105 of constructing sub-images S ₁ (x, y) , S ₂ (x, y) , S ₃ (x, y) by constructing an area-

With respect to the third differentiation sub-images each of A _1, A _2, and the error filtered by the A ₃ values to produce the _{R 1 (x, y),} R 2 (x, y), R 3 (x, y) [107]

Converting the R ₂ (x, y) , R ₃ (x, y) image to a 4-gray value [109]

A merging step 111 for merging the transformed image of the generated image R ₁ (x, y) of the error filtering step 107 with the transformed image of the 4 gray transforming step 109 to output the final optimized image G (x, y) .

The step of generating a gray image [101] among the above steps may be implemented by an algorithm as shown in FIG. In addition, the partition resetting step [103] may be implemented by an algorithm configured as shown in FIG.

In the above steps, a step 105 of constructing a sub-image by composing a region-by-region image with three overlapping regions from a divided region (hereinafter referred to as "overlapping region 3 differentiation step") will be described. The resultant image is determined by expressing a value with a high frequency density among the 256 steps shown in the original image as a main four-step intensity value. In this case, the 252 step intensity step processing among the four intensity values remains as a problem. This is because the processing of the lossy intensity step values determines the quality of the resulting image. A preprocessing process is performed on the image in such a way as to maximize the intensity change step of the original image while reducing the loss rate as much as possible.

The overlapping region 3 differentiation step [105] is a process of dividing the input image into three separate images (sub-images), and based on the values to be selected as the four intensities in the resultant image. One intensity intensity representative value is selected, and the area between the representative values of the next stage is made into one three-divided image based on the representative intensity value. The three triad images commonly include areas of up to the next representative lightness value and the next lightness value to be selected, and these common areas are portions that indicate the intensity step change from the error diffusion filtering to the next step . This redundant area tripletization step [105] can be implemented by an algorithm as shown in Figs. 1D and 2E.

There is a process of converting to 4 gray values [109] after error diffusion dithering (filtering) is applied to three separate sub-image sources that have undergone redundant area triangulation, and this 4 gray conversion step [109] 1f. ≪ / RTI >

The merging step of merging the image R ₁ (x, y) generated in the error filtering step 107 and the image converted in the 4-gray conversion step 109 to output the final result image G (x, y) [111] can be implemented by the algorithm shown in FIG. 1G. At this time, the resulting image is represented by the density of the points made up of the four intensity values for the middle stage of the four-level gray value.

(2) Extension of image characterization dithering for a 1-bit monochrome display (see Figs. 2a-8)

The present embodiment is a method developed in the image characterizing dithering process for a 1-bit monochrome display device. First, an image characterization dithering process for a 1-bit monochrome display device is briefly described. It consists of steps of applying first high-pass filtering to the input image, selecting the second best result, separating the third channel, and fourth error diffusion dithering. can do.

As an extension thereof, the 4-way characterizing dithering method according to the present invention is characterized in that, instead of the error diffusion dithering, which is the last step in the binarization method for a 1-bit display device as shown in FIG. 2B, This dithering, which performs thresholding with a threshold value, is applied. The overall configuration will be described with reference to FIG.

FIG. 2B shows an overall process of characterizing dithering according to the present invention.

First, color separation is performed for each channel in order to take advantage of characteristics of each color of an image [201]. That is, I _{r, g, and bi} (x, y) are generated by separating the original color image F (x, y) into red (R), green (G), blue (B), and intensity do. FIG. 2B shows an algorithm for implementing the channel separation step [301].

High pass filtering (HPF) is then performed on each channel separated image. High-pass filtering is a step for characterizing a boundary portion of an image to emphasize a color portion. At this time, a 3 × 3 spatial mask filter (described below) is used. An algorithm for implementing high-pass filtering is shown in FIG.

After performing the high-pass filtering, a step (Detect A []) [205] of resetting the segmented region based on the distribution of colors in the filtered image and a step of constructing an image of region by three overlapping segments from the segmented region sub-image _{S 1 (x, y),} S 2 (x, y), S 3 (x, y) step [207] and, with respect to the third differentiation sub-images each of a _1, a _2, a to establish a _three- by a thread when held by the value _{R 1 (x, y),} R 2 (x, y), R 3 and step [209] to generate the _{(x, y), R 2} (x, y) and R ₃ (x, y) of the 4-gray transform step [109] and the generated image R ₁ (x, y) of the thresholding step [209] And then outputting the final image G (x, y) .

Among the above steps, the step of resetting the divided area [205] may be implemented by an algorithm configured as shown in FIG. 1C as described above.

The step of constructing the sub-image by constructing the region-by-region image from the divided region into three overlapping regions is the same as that described in the first embodiment above. That is, the redundant region 3 differentiation step [207] divides the input image into three separate images (sub-images), and the four intensities are selected based on the resultant image. One intensity intensity representative value is selected, and the area between the representative values of the next stage is made into one three-divided image based on the representative intensity value. The three triad images commonly include regions to be selected from the representative lightness values to the next lightness values, which are the portions where the common regions will exhibit the intensity step change to the next step through the thresholding [209] . This redundant area tripletization step [207] may be implemented by an algorithm as shown in Figs. 1D and 2E.

The step of performing thresholding 209 on three separate sub-image sources that have undergone the redundant area 3 differentiation process can be implemented by the algorithm shown in FIG. 2D, and the process of converting to the next 4 gray values [211 Can be implemented by the same algorithm as in Fig. 1F as in the first embodiment.

(X, y) generated by merging the image R ₁ (x, y) generated in the thresholding step 209 and the image converted in the 4 gray conversion step 211 to output the final result image G Step [213] can be implemented by an algorithm as shown in FIG. At this time, the resulting image is represented by the density of the points made up of the four intensity values for the middle stage of the four-level gray value.

The result that the color image is actually converted by the present invention is shown in Fig. 8 shows that the color image is optimized for display on the 4 gray display device. When printing on paper with a monochrome laser printer, the original image may be displayed with the best quality due to the high density output of the printer, but the result should be compared considering that the personal portable terminal does not have the high resolution density such as the printer.

As in the conventional art, the simple binary-coding result image does not faithfully reproduce the original tone because it converts only the tone of the original image into black and white, and the image according to the conventional dithering result does not represent the halftone The lines are not accurately expressed and the images are unclear as a whole. On the other hand, the resultant image according to the present invention is natural in color representation, and the outline of the image is clearly expressed, so that the original image can be faithfully displayed on the personal mobile phone .

Specific description of the high-pass filtering step

Let us now consider the spatial filter used in the high-pass filtering step [203] from the description of the second embodiment. The use of a spatial mask for image processing is called spatial filtering, and the masks at this time are called spatial filters. The spatial mask is a matrix in which the spatial domain is a unit. The spatial domain is a sub-image region of a square region formed by neighboring pixels centered at a certain pixel (x, y) of the input image.

FIG. 3A shows a 3 × 3 neighboring pixel for any pixel (x, y), and FIG. 3B shows a typical 3 × 3 mask. If the center of this mask is on one point (x, y) of the input image and the neighboring pixels are z ₁ , z ₂ , ..., z ₉ , the intensity step result at point (x, y) Expressed as the expression R.

The spatial filtering is completed by applying the mask from the first pixel to the last pixel of the input image and repeating the same calculation process.

FIG. 4 shows a basic highpass spatial filter, which is a conventional 3 × 3 sharpening filter. This is a method of obtaining a high-pass filtered result in addition to a method using a high-pass filter, and can be obtained by a difference between an input image and a low-pass filtered image.

In other words, .

Here, " high-boost " filtering can be performed by multiplying the original image by the weight A. In other words,

When A = 1 in the above equation, it is the same as the result of standard high-pass filtering. If A> 1 , the result is closer to the original image, and the A value changes the degree of edge enhancement. Here, the process is the same as the use of the mask shown in FIG. 5 with the following w (weight) value as a median value.

Methods for high-pass filtering are more suitably represented in the frequency domain. For the frequency domain processing, the original image data is transformed into the frequency domain by FFT (Fast Fourier Transform), then filtered by the filter mask, and then the result image is obtained by inverse FFT again.

The ideal high-pass filter in a two-dimensional space satisfies the following relation. As shown in FIG. 6, D ₀ is a cutoff frequency at which H (u, v) transitions from zero to one.

The actually used filter uses a Butterworth filter which is more suitable for the real world with a gentle frequency response than IHPF as shown in FIG. Butterworth high-pass filter (BHPF) is when n have the difference is given by the following relation, n = 1 when equal to 7 il.

When D (u, v) = D _0, H (u, v) has a maximum value of 1/2. Otherwise, Is used as the maximum value of H (u, v) . Therefore, the above equation can be modified as follows.

Implementation with image conversion program

The color image optimization method and the algorithm for the 4 gray display device according to the present invention described above can be implemented by a program in a predetermined programming language and implemented by a program. It is obvious that the computer recording medium (CD, PROFIBUS, hard disk, ROM, etc.) storing the program thus implemented is included in the technical scope of the present invention.

The image conversion method according to the present invention is applicable to all fields aimed at obtaining an optimized image for a 4 gray display device from a color image. Especially, it can be applied to the development of software for making 4 gray image optimized for LCD screen of personal mobile phone in connection with wireless Internet use. This method has been developed in order to simplify and simplify the generation of image information having a substantial value suitable for a wireless Internet environment using a personal mobile phone as a terminal.

As a few examples, when creating a wireless internet homepage, when trying to transform a company logo or a simple picture without developing a new design, when trying to create a 4 gray image by maximizing existing color image for game and advertisement, When it is desired to represent individual photographs and the like appropriately on a personal mobile phone or the like and to express a color image which is used in the past in a machine device having a liquid crystal display device which displays 4 gray images such as a PDA having a 4 gray display device, Can be applied.

Claims (4)

A method for generating an optimized result using color information of an original image to output an optimized image for displaying on a 4 gray display device from a color image, Generating a gray image from the input color image, Resetting the divided area based on the distribution of colors in the gray image, Constructing a sub-image S ₁ (x, y) , S ₂ (x, y) , S ₃ (x, y) by constructing a region-by-region image from three sub- With respect to the third differentiation sub-images each of A _1, A _2, and the error filtered by the A ₃ values to produce the _{R 1 (x, y),} R 2 (x, y), R 3 (x, y) , Converting the R ₂ (x, y) and R ₃ (x, y) images to 4 gray values, And a merging step of merging the generated image R ₁ (x, y) of the error filtering step and the converted image of the 4 gray conversion step to output a final 4 gray image G (x, y) Optimization of Color Image for. A method for generating an optimized result using color information of an original image to output an optimized image for displaying on a 4 gray display device from a color image, A step of color-separating each channel to convert the image into color, Performing high pass filtering (HPF) on each channel separated image, Resetting the segmented region based on the distribution of colors in the filtered image, Constructing a sub-image S ₁ (x, y) , S ₂ (x, y) , S ₃ (x, y) by constructing a region-by-region image from three sub- With respect to the third differentiation sub-images each of A _1, A _2, and the thread when held by the A ₃ values to generate _{R 1 (x, y),} R 2 (x, y), R 3 (x, y) step, Converting the R ₂ (x, y) and R ₃ (x, y) images to 4 gray values, And a merging step of merging the generated image R ₁ (x, y) of the thresholding step with the converted image of the 4-gray conversion step to output a final binary image G (x, y) A method for optimizing color images for a device. The method of claim 2, wherein the high-pass filtering is implemented using a 3 × 3 spatial mask filter. A computerized recording medium recording a method of optimizing a color image for a 4-gray display device, which is configured as in claim 1 or 2 or 3.