JP4770124B2 - Image display method - Google Patents

Description

  The present invention relates to an image display method for displaying an image close to an original image by increasing the gradation display in a display device such as a plasma display panel (PDP).

Conventionally, an error diffusion method is known as an image display method for artificially increasing the number of gradations of a display device (see, for example, Non-Patent Document 1). FIG. 7 is a diagram for explaining an example of processing of this error diffusion method. The circles in FIG. 7A schematically show the pixels constituting the image. The input image data of M gradation corresponding to the pixel of interest indicated by the shaded circle is converted into image data of N gradation having a smaller number of gradations, and this is used as display data for the pixel of interest. However, at that time, an error occurs between the input image data and the display data. Therefore this error, four of the diffusion pixels indicated by black circles in the input image signal with respect to (the pixel A to the right, and the pixel B of the lower right, and the pixel C of the true underlying lower left pixel D), 7 Divide by 1: 5: 3, then spread and add. As a result, for example, in the pixel A, display data is obtained by converting input image data of M gradation obtained by adding the diffused error and an input image signal to itself into image data of N gradation. Also at this time, an error between the input image data of M gradation and the display data is diffused to surrounding pixels. The above processing is repeated while sequentially setting the pixel to which the input image signal is input as the target pixel.

  A circuit for performing the above-described error diffusion method is shown in FIG. In FIG. 7B, reference numeral 700 is an input image signal for the target pixel, reference numeral 701 is an adder that adds an error diffused from surrounding pixels to the input image signal, and reference numeral 702 is an adder from the adder 701. The input image data of M gradation, which is the output of the input image signal and the diffused error, are converted into N gradation image data to be display data, and an error is generated from the display data and the input image data. Is a display data / error divider that calculates Reference numerals 703 to 706 denote delay units for adjusting timing in order to diffuse errors to desired pixels (T denotes a delay of one data period, H denotes a delay of one scanning period), and reference numerals 707 to 710 It is a multiplier for giving a predetermined weight to the error. In general, the multiplier 1 is 7/16 times, the multiplier 2 is 3/16 times, the multiplier 3 is 5/16 times, and the multiplier 4 is 1/16 times.

  Here, in the error diffusion method described above, a specific pattern generated when an error is diffused is observed, which may cause a problem that image quality deteriorates. In view of this, an error diffusion method has been proposed in which an error generated in a pixel of interest is diffused to a pixel at the same position as the pixel of interest in the next field (see, for example, Patent Document 1). FIG. 8 is a schematic diagram for explaining the error diffusion method, and shows a state in which three field images that are temporally continuous centering on the current field image are arranged. The pixels constituting the field image are schematically shown by circles. An error occurring in a pixel of interest indicated by a shaded circle in the current field (field of interest) is indicated by a black circle in pixels A, B, C, D, and the next field in the current field. The error is diffused to the pixel E at the same position as the target pixel.

  A circuit for performing this processing is as shown in FIG. For the error diffusion circuit shown in FIG. 7B, an error divider 900 that divides the generated error into the current field and the next field, and a timing for diffusing the error generated in the target pixel to the pixel E. Delay unit 901 (F indicates a delay of one field period) and a multiplier 8 (reference numeral 902) for adding a predetermined weight to an error diffused to the pixel E.

In other words, the error diffusion process described above modulates the error between the input image data and the display data into a high-frequency component, and spatially finely distributes the false contour generated by the reduction in the number of gradations. It is also a process to reduce the visual influence.
R. W. Floyd and L. Steinberg “An Adaptive Algorithm for Spatial Gray Scale” SID75 Digest (1976) JP-A-6-118920

  When the input image data is three-dimensional data having three axes of horizontal (x), vertical (y), and time (field, t), the image display method according to the configuration of FIG. This is equivalent to passing the error component through a filter having frequency characteristics shown in FIG. 10A shows the frequency characteristic in the horizontal (x) -vertical (y) plane, FIG. 10B shows the frequency characteristic in the horizontal (x) -time (t) plane, and FIG. ) Indicates frequency characteristics in the vertical (y) -time (t) plane. FIG. 10A shows that the error component after the error diffusion process has a high spatial frequency in the field. This indicates that errors generated when the input M gradation image data is converted to N gradations and displayed are less likely to be observed due to fine dispersion in the field. In contrast, FIGS. 10B and 10C show that the frequency characteristics are flat in the time direction. This means the following: That is, in general, the error component usually varies from field to field due to the influence of noise or the like included in the original image. Since the frequency characteristics shown in FIGS. 10B and 10C are flat in the time direction (field direction), the error component fluctuates at a time frequency comparable to the noise included in the original image. . Since this is often a relatively gradual change, the pattern resulting from error diffusion is observed as a fluctuating noise, and the image quality deteriorates.

  The image display method having the configuration shown in FIG. 9 is equivalent to passing the error component through a filter having frequency characteristics shown in FIG. FIG. 11A shows the frequency characteristics in the horizontal-vertical plane, FIG. 11B shows the frequency characteristics in the horizontal-time plane, and FIG. 11C shows the frequency characteristics in the vertical-time plane. 11 (b) and 11 (c), it can be seen that the low-frequency components are slightly removed in the time direction, but there are still many low-frequency components, and thus the patterns caused by error diffusion are fluctuating. It will be observed as moving noise, and the image quality will deteriorate.

  The present invention has been made to solve such problems, and provides an image display method that improves the image quality by changing the pattern caused by error diffusion at high speed for each field and making it difficult to observe. Objective.

In order to achieve the above object, the image display method of the present invention converts M-gradation input image data for a pixel of interest into N-gradation (M> N) image data, which is used as display data for the pixel of interest, By multiplying an error between the input image data and the image data by a predetermined diffusion coefficient and diffusing and adding to the input image signal for the pixels in the field of interest and the next field, the respective input image data are obtained. An image display method for expressing input image data of M gradation for a target pixel by converting it to display data of N gradation (M> N), wherein the input image data is expressed in three dimensions of horizontal, vertical, and time. in the one configured, horizontal of the target field - in a vertical plane, the right of the target pixel, lower right, beneath and lower left pixel of each pixel a, pixel B, pixel C and pixel In the horizontal-vertical plane of the next field, a pixel at the same position as the target pixel is defined as a pixel E, and pixels immediately above, directly below, right adjacent and left adjacent to the pixel E are defined as a pixel F, a pixel G, Let pixel H and pixel I be the pixel A, pixel H, pixel E, and pixel I, respectively, the pixel adjacent to the right, lower right, directly below, and lower left of the pixel of interest in the horizontal-time plane, and vertical-time When the pixel adjacent to the right side, lower right, right below, and lower left in the plane is the pixel C, the pixel G, the pixel E, and the pixel F, respectively, the horizontal-vertical plane and the horizontal-time plane In each plane in the vertical-time plane, the error multiplied by a predetermined diffusion coefficient is diffused to the pixels adjacent to the right side of the target pixel, the lower right side, the lower right side, and the lower left side. Next, lower right, directly below and lower left When the diffusion coefficients of diffusion to each pixel are K1, K2, K3, and K4, the magnitude relationships among K1, K2, K3, and K4 are horizontal-vertical plane, horizontal-time plane, and vertical-time. It is the same in each plane within the plane .

  According to the image display method and the image processing apparatus of the present invention, the pattern caused by error diffusion is not easily observed because the pattern fluctuates at a high speed for each field, and the image quality can be improved.

According to the first aspect of the present invention, the input image data of M gradation for the target pixel is converted into image data of N gradation (M> N), which is used as display data for the target pixel, and the input image The pixel of interest is obtained by diffusing and adding an error between the data and the image data to the input image signal for the pixels in the field of interest and the next field to obtain the respective input image data. An image display method for expressing input image data of M gradations by converting them into display data of N gradations (M> N), wherein the input image data is configured in three dimensions of horizontal, vertical, and time. ones and to the horizontal of the target field - and in the vertical plane, the right of the target pixel, lower right, beneath and lower left pixel of each pixel a, pixel B, a pixel C and the pixel D, the next full In the horizontal-vertical plane of the field, a pixel at the same position as the target pixel is a pixel E, and pixels immediately above, directly below, right adjacent and left adjacent to the pixel E are a pixel F, a pixel G, a pixel H, and a pixel, respectively. Let I be the pixel A, the pixel H, the pixel E, and the pixel I, respectively, in the horizontal-time plane, right next to the pixel of interest, lower right, right below, and lower left, respectively. Let the pixel C, the pixel G, the pixel E, and the pixel F be the pixel adjacent to the right side, the lower right side, the lower right side, and the lower left side of the pixel of interest, respectively. Within each plane, a product of the error multiplied by a predetermined diffusion coefficient is diffused to the right next to the target pixel, the lower right, directly below, and the lower left pixel, and the right next to the target pixel, the lower right , Diffusion to each pixel below and below left Assuming that the diffusion coefficients are K1, K2, K3, and K4, the magnitude relationships among the K1, K2, K3, and K4 are within horizontal-vertical planes, horizontal-time planes, and vertical-time planes. This is the same image display method.

The invention of claim 2 is the invention according to claim 1, by the predetermined diffusion coefficients, half or more of the error between the pixel A and the pixel B and the pixel C diffused into said pixels D, are added, also characterized in that less than half of the error is diffused to said pixel I and the pixel E and the pixel F and the pixel G and the pixel H, is added Is.

Further, the invention according to claim 3 of the present invention, in the invention described in claim 1, by the predetermined diffusion coefficient, 3/4 of the error between the pixel A and the pixel B to the pixel C the Each pixel D is divided and diffused and added at a ratio of 7: 1: 5: 3, and 1/4 of the error is the pixel E, the pixel F, the pixel G, the pixel H, and the pixel I. And divided and added at a ratio of 6: 4: 1: 1: 4.

  According to the present invention, a pattern caused by error diffusion is changed at a high speed for each field so that the pattern is not observed. Therefore, as in the horizontal-vertical plane, the high-frequency component of the error component in the horizontal-time plane and the vertical-time plane is large and the low-frequency component is small. This is based on the idea that the shape of the frequency characteristic in the horizontal-vertical plane in the conventional error diffusion method described with reference to FIG. 10 is also reflected in the time direction.

  Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

FIG. 1 is a diagram for explaining an outline of error diffusion processing in an image display method according to an embodiment of the present invention. Three consecutive field images centering on an image in a current field (field of interest) are shown. It shows how they are lined up. The circles schematically show the pixels constituting the image of each field, and the shaded circles in the current field are the target pixels. Further, as shown by black circles in the current field, the pixel of interest, the pixel (pixel A) to the right, the lower right pixel (pixel B), the true lower pixel (pixel C), the lower left pixel (pixel D ), And a pixel in the same position as the pixel of interest (pixel E), a pixel directly above (pixel F), a pixel immediately below (pixel G), a pixel right next (pixel H), indicated by a black circle in the next field, The pixel on the left (pixel I) is a diffused pixel that diffuses an error generated in the target pixel. In the above, the generated error is divided into the current field and the next field and diffused. However, if many errors are diffused in the next field, the effect of the error diffusion processing in the current field is weakened and the multi-gradation is reduced. It becomes impossible to express. Therefore, it is desirable that more than 1/2 of the total error is diffused in the current field. In the following description, it is assumed that 3/4 of the error generated in the target pixel is diffused in the current field, and 1/4 is diffused in the next field. Also, in the current field, the error is divided, diffused, and added to the ratio of 7: 1: 5: 3 for each of the pixel A, pixel B, pixel C, and pixel D, and in the next field, The error is divided into a ratio of 6: 4: 1: 1: 4 for each of the pixel E, the pixel F, the pixel G, the pixel H, and the pixel I, diffused, and added.

  2 to 4 are diagrams for explaining error diffusion in each of the horizontal-vertical plane, the horizontal-time plane, and the vertical-time plane.

  In the image display method according to the present embodiment, the error diffusion direction in the horizontal-vertical plane shown in the shaded plane in FIG. 2A is as shown in FIG. The ratio of the magnitude of the error is pixel A: pixel B: pixel C: pixel D = 7: 1: 5: 3.

  Also, the error diffusion direction in the horizontal-time plane shown in the shaded plane in FIG. 3A is as shown in FIG. 3B, and the ratio of the magnitude of the diffused error is Pixel A: Pixel H: Pixel E: Pixel I = 21: 1: 6: 4.

  Also, the error diffusion direction in the vertical-time plane shown in the shaded plane in FIG. 4A is as shown in FIG. 4B, and the ratio of the magnitude of the diffused error is Pixel C: Pixel G: Pixel E: Pixel F = 15: 1: 6: 4.

  From the above, in any plane, the error diffusion direction is the same, the diffusion coefficient to the right is the largest, the downward diffusion coefficient is the second, the lower left diffusion coefficient is the third, the right The downward diffusion coefficient is minimized.

  The error diffusion method as described above is equivalent to passing an error component through a filter having frequency characteristics shown in FIG. 5A shows frequency characteristics in the horizontal (x) -vertical (y) plane, FIG. 5B shows frequency characteristics in the horizontal (x) -time (t) plane, and FIG. 5C shows the frequency characteristics. The frequency characteristics in the vertical (y) -time (t) plane are shown. FIG. 5A shows that the error component after the error diffusion process has a high spatial frequency in the field. This indicates that errors that occur when the input image data of M gradations are converted into display data of N gradations and expressed are less likely to be observed due to fine dispersion in the field. 5B and 5C show that the error component after signal processing of the present invention has a low frequency component and a high frequency component in the time direction. This means that the generated error fluctuates at a high speed in the field direction, and it is difficult to observe a pattern caused by error diffusion, and as a result, the image quality is improved.

  As described above, the error diffusion process is individually considered for each of the horizontal-vertical plane, the horizontal-time plane, and the vertical-time plane, and the diffusion direction and coefficient are matched to the horizontal-vertical plane having ideal frequency characteristics. Thus, it is possible to make the frequency characteristics in each plane close to an ideal shape.

  FIG. 6 is a block diagram showing a schematic configuration of a circuit for realizing the above-described image display method. Compared to the configuration shown in FIG. 9, multipliers 5 to 7 (reference numerals 600 to 602) and multiplier 9 (reference numerals). 603) and delay devices 604 to 607 are added. In the delay units 703 to 706, 604 to 607, and 901, T represents a delay of one data period, H represents a delay of one scanning period, and F represents a delay of one field period. In addition, the error divider 900 passes 3/4 of the total error to the error diffused in the current field (path through the multipliers 1 to 4), and the error diffused in the next field (multipliers 5 to 9). Route) is assigned 1/4 of the total error. The value of the delay device 901 is FH. The multipliers 5 to 9 are 4/16, 1/16, 6/16, 4/16, and 1/16, respectively.

  In the above description, an example has been shown in which 3/4 of the total error is diffused in the current field and 1/4 is diffused in the next field. However, the present invention is not limited to this, and horizontal- Similar effects can be obtained by providing frequency characteristics such that there are few low-frequency components and many high-frequency components in any of the vertical, horizontal-time, and vertical-time planes.

  Further, the present invention is not limited to the diffusion direction as described above, and there are few low frequency components and many high frequency components in any of the horizontal-vertical, horizontal-time, and vertical-time planes. By having frequency characteristics, the same effect can be obtained.

  Note that the present invention is not limited to the diffusion coefficient as described above, and there are few low frequency components and many high frequency components in any of the horizontal-vertical, horizontal-time, and vertical-time planes. By having frequency characteristics, the same effect can be obtained.

  As described above, according to the image display method according to the embodiment of the present invention, in the error diffusion processing method in the conventional image display method, the error is modulated at a high frequency in the horizontal-vertical plane. Since the direction included a low-frequency component, the pattern due to error diffusion was observed to move slowly and slowly from field to field. In addition, since it can be modulated to a high frequency even in the vertical-time plane, the pattern resulting from error diffusion varies at a high speed for each field, and the pattern is difficult to be observed, thereby improving the image quality.

  The image display method according to the present invention has an effect that a smooth pseudo gradation can be expressed without observing a pattern caused by error diffusion even in a display device having a smaller number of display gradations than input image data. This is useful as an image display method for matrix displays such as PDP and PDP.

The figure for demonstrating the outline | summary of the error diffusion process in the image display method by one embodiment of this invention The figure for demonstrating the error diffusion in the horizontal-vertical surface in the image display method by one embodiment of this invention The figure for demonstrating the error diffusion in the horizontal-time surface in the image display method by one embodiment of this invention The figure for demonstrating the error diffusion in the perpendicular | vertical-time surface in the image display method by one embodiment of this invention. The figure which shows the frequency characteristic of the filter equivalent to the error diffusion method in the image display method by one embodiment of this invention The block diagram which shows schematic structure of the circuit of the error diffusion process in the image display method by one embodiment of this invention The figure for demonstrating an example of the process of the error diffusion method in the conventional image display method The figure for demonstrating an example of the process of the error diffusion method in the conventional image display method FIG. 8 is a block diagram showing a schematic configuration of a circuit for error diffusion processing in the conventional image display method shown in FIG. The figure which shows the frequency characteristic of the filter equivalent to the error diffusion method in the conventional image display method shown in FIG. The figure which shows the frequency characteristic of the filter equivalent to the error diffusion method in the conventional image display method shown in FIG.

600 Multiplier 5
601 Multiplier 6
602 Multiplier 7
603 Multiplier 9
604 delay device 605 delay device 606 delay device 607 delay device 700 input image data 701 adder 702 display data / error divider 703 delay device 704 delay device 705 delay device 706 delay device 707 multiplier 1
708 Multiplier 2
709 Multiplier 3
710 Multiplier 4
900 Error divider 901 Delay device 902 Multiplier 8

Claims (3)

The input image data of M gradation for the target pixel is converted into image data of N gradation (M> N), and this is used as display data for the target pixel.
By multiplying an error between the input image data and the image data by a predetermined diffusion coefficient and diffusing and adding to the input image signal for the pixels in the field of interest and the next field, the respective input image data are obtained. ,
An image display method for converting and expressing input image data of M gradation for a target pixel into display data of N gradation (M> N),
The input image data is configured in three dimensions, horizontal, vertical, and time ,
In the horizontal-vertical plane of the field of interest, the pixels adjacent to the right, lower right, directly below, and lower left of the pixel of interest are pixel A, pixel B, pixel C, and pixel D, respectively, and the horizontal-vertical plane of the next field , The pixel at the same position as the target pixel is a pixel E, and the pixels immediately above, directly below, right adjacent and left adjacent to the pixel E are a pixel F, a pixel G, a pixel H, and a pixel I, respectively. The pixel A, the pixel H, the pixel E, and the pixel I are the pixel adjacent to the right side, the lower right, directly below, and the lower left of the pixel of interest, respectively, and within the vertical-time plane, When the lower, right-bottom and lower-left pixels are the pixel C, the pixel G, the pixel E, and the pixel F, respectively,
In each of the horizontal-vertical plane, the horizontal-time plane, and the vertical-time plane, the error multiplied by a predetermined diffusion coefficient is the right next to the target pixel, the lower right, the lower right, and the lower left. If the diffusion coefficients of diffusion to each pixel and diffusion to each pixel on the right side, lower right, right below and lower left of the pixel of interest are K1, K2, K3, K4, respectively, K1, K2, K3, K4 The image display method characterized in that the magnitude relationship between them is the same in each of the horizontal-vertical plane, the horizontal-time plane, and the vertical-time plane . Due to the predetermined diffusion coefficient, ½ or more of the error is diffused and added to the pixel A, the pixel B, the pixel C, and the pixel D, and ½ or less of the error is the pixel E. the image display method according to claim 1, wherein the said pixel F and the pixel G and the pixel H diffused to the pixel I, are added. According to the predetermined diffusion coefficient, 3/4 of the error is diffused and added to the pixel A, the pixel B, the pixel C, and the pixel D at a ratio of 7: 1: 5: 3, respectively. Further, 1/4 of the error is diffused and added to the pixel E, the pixel F, the pixel G, the pixel H, and the pixel I in a ratio of 6: 4: 1: 1: 4, respectively. The image display method according to claim 1 .