KR20010080280A - Addressing method for plasma display panel based on separate even-numbered and odd-numbered line addressing - Google Patents

Abstract

The method of the present invention is performed according to whether different coding of a column control word is related to the even or odd lines of the word, the difference being that at least m consecutive bits of a particular rank are each from one control word to another control word. With different weights, the weighted sum of the bits remain equally from one control word to another, thereby obtaining substantially different write instants from one line to the next.

The present invention is applied to the addressing of plasma panels.

Description

ADDRESSING METHOD FOR PLASMA DISPLAY PANEL BASED ON SEPARATE EVEN-NUMBERED AND ODD-NUMBERED LINE ADDRESSING}

On a plasma screen, gray levels are not generated in a conventional manner using amplitude modulation of the signal, but rather using temporal modulation of the signal by exciting the corresponding pixels for larger or smaller times depending on the desired level. . It is the integration phenomenon by the eye that makes it possible to provide that gray level. The integration is performed during the frame scan time.

Since the eye actually integrates much faster than the frame duration, it is easy to notice variations in levels that do not reflect realism in the case of certain transitions of addressing bits. Thus, contour defects or "contouring" as known may appear in the movie. The defect can be compared to poor temporal recovery of gray levels. More generally, a false color appears in the outline of the object, with each cell having a color component possibly affected by the phenomenon. The phenomenon is more harmful when it occurs in a relatively homogeneous zone.

A simple theoretical solution for limiting the appearance of false contours is known from the prior art described in the French patent application, for example, filed April 25, 1997 and published under publication number FR 2762704, the solution of said French patent application. Is to increase the number of sub-scans so that disturbances related to video level adjustment from one frame to another are minimized. The additional sub-scans needed result from the scan being saved by the simultaneous addressing of two adjacent lines. However, the simultaneous addressing causes a loss of resolution, i.e., the possibility of code redundancy, resulting in the loss of information copied from one line to another line obtained by recording gray levels. However, it is not possible to suppress the magnitude of the resolution loss.

Another problem with the prior art relates to priming conditions.

One of the features of a plasma cell is that the cell has a triggering threshold that is affected by the state of its immediate neighbor. The cell will be more easily excited when adjacent cells are excited, which actually represents a priming phenomenon. Since the barriers separating the various cells are not completely enclosed, a certain number of free electrons from the excited adjacent cells will facilitate the excitation of the addressed cell.

The priming problem is, in fact, amplified by the nonuniformity of the panel. In order to facilitate the excitation of the cell, it is always possible to adjust the control voltage, but this becomes impossible when the glass plates do not have the same spacing across the entire panel, for example. In that case, the compromise set at the level of its control voltage makes it impossible to optimize the light emission of all cells.

Another problem with the prior art involves low levels of quantization.

Plasma panels, unlike cathode ray tubes, have a linear response, ie the luminance level emitted is exactly proportional to the video level. Today's display systems are mostly based on the use of cathode ray tubes. Next, a priority compensation operation on the response of the cathode ray tube is performed at the image-acquisition level. It is therefore necessary to perform inverse correction (gamma correction) in order to eventually obtain the actual information so that the signal can be displayed accurately on the plasma panel.

Figure 1 shows the profile of the compensation curve 1 for the response of the cathode ray tube on emission, with the abscissa axis representing the input video level and the ordinate axis representing the output video level after correction. Curve 2 corresponds to a linear response obtained after applying a correction as shown in FIG. 3.

As a result of the correction, the low level quantization is very limited as long as some level of the input signal can correspond to the level of the output signal. This is especially true for low levels in the region demarcated by 4, which corresponds to a single output level where the input level between 0 and 15, for example, is zero.

In order to provide the low level of completeness, it will be necessary to use quantization bits greater than 8 (eg 10 or 12).

The present invention relates to an addressing method and apparatus for a plasma panel based on separate addressing of even and odd lines.

1 shows a compensation curve for a response curve of a cathode ray tube.

2 is a timing diagram illustrating coding levels as a function of time.

3 is a scanning configuration diagram for a plasma panel according to the prior art.

4 is a scanning configuration diagram for a plasma panel according to the present invention.

5 is a timing diagram for writing two consecutive lines in accordance with the present invention, for bits of a column control word having different weights.

6 is a timing diagram for writing two consecutive lines in accordance with the present invention, for bits of a column control word having the same weight.

7 shows an example of writing on two consecutive lines for bits of a column control word having the same weight.

8 shows an example of writing on two consecutive lines for bits of a column control word having different weights.

9 shows a device according to the invention.

It is an object of the present invention to address the drawbacks mentioned. Accordingly, the subject matter of the present invention is a method for addressing cells arranged in a matrix array,

Each cell is located at the intersection of lines and columns, and the array has line inputs and column inputs for displaying gray levels defined by video words for forming a digital video signal and confining an image, each of which has its own input. Receives a control word for that column that corresponds to a video word for a line addressed for that column, and that word that is transmitted continuously consists of n bits, each sequence corresponds to a sub-scan, and each bit in the word The method, with or without triggering, in response to its state, i.e., light emission of a cell in an addressed line and column receiving a control word, for a time proportional to the weight of the bit.

Other coding of a column control word is performed depending on whether it is related to the even or odd lines of the word, the difference being that at least m consecutive specific rank bits have different weights from one control word to another control word, The sum of the weights of the bits is such that it remains the same from one control word to the other control word, thereby obtaining substantially different write instants from one line to the next.

According to a variant of the method, the writing is simultaneous on two consecutive lines for at least the first bit of the m consecutive bits of the control word for one of the two lines.

According to another variant, at least two consecutive lines are selected simultaneously for at least one of the bits of the column control word, the weights of which bits being the same from one control word to the other.

According to another variant, at least one of the bits having the same weight from one control word to the other control word is used to code a partial luminance value common to two consecutive lines, and the recording relates to one of the two lines. It is concurrent on the line for that bit of the control word.

According to another variant, the method is implemented for a limited number of matrix array lines, the lines corresponding to regions of the image defined by the video signal with strong vertical transitions, with the other regions being equal to all of the column control words. A sub-scan is used that corresponds to the addressing method with the same weight from line to another line.

According to another variant, the method is implemented for images with strong vertical transitions, and the other image uses an addressing method in which all of the column control words have the same weight from one line to another.

According to another variant, from the first addressing method comprising n sub-scans, a larger number of bits containing a larger number of sub-scans and the column control word having the same weight from one line to another line are taken. Switching to the second addressing method has a selection of the line (l) with writing the bits having different weights on the line (1) in the first method. This is done by replacing the line 1 and the preceding or subsequent line for simultaneous writing on both lines, either immediately preceding the line or immediately following the line.

The invention also relates to an apparatus for implementing the above-described method,

Video processing circuitry for processing received video data,

A corresponding memory for transcoding the data;

A video memory for storing transcoded data, comprising: a video memory linked to a heat supply circuit for controlling column addressing of a plasma panel based on a column control word;

The apparatus comprising a control circuit for a line supply circuit, linked to a video processing circuit for selecting a line,

The video processing circuit and the transcoding circuit perform different coding of the column control word depending on whether the word is related to even or odd lines, the difference being at least m consecutive specific rank bits of the bits to be transmitted. Has different weights from one control word to another, and the sum of the weighted weights of bits remains the same from one control word to another, thereby obtaining substantially different write instants from one line to the next. It is characterized by being.

According to a variant embodiment, the apparatus is characterized in that the circuit for controlling the line supply circuit simultaneously transmits two consecutive lines while the first bit of the consecutive bits of the control word relating to one of the two lines is transmitted by the column supply circuit. It is characterized by selecting.

According to another variant, the apparatus controls the column corresponding to addressing according to n sub-scans or addressing corresponding to a larger number of sub-scans, as a function of the luminance variation from one line to another line of the image. And a selection circuit for receiving the video data to select the coding of the word.

The addressing method according to the invention is to separate the addressing of even lines from the addressing of odd lines by using different coding of the column control words. The moments of writing from one line to another for a particular bit of control word are substantially different. Thus, priming of the excitation of the cell is facilitated.

The method makes it possible to perform partial and variable copying of video information from one line to another. Thus, it is possible to deal with the tradeoff of the number of sub-scans / vertical resolution loss. Next, depending on the content function of the video, it is possible to adjust the number of sub-scans for each of the line pairs and, accordingly, to adjust the maximum difference allowed between the two luminance values, which allows for errors smaller than the LSB. .

Through the present invention, the contour effect is eliminated or at least greatly reduced, and low level quantization is improved.

Other features and advantages of the invention will be apparent from the following description of the accompanying drawings, which are provided through non-limiting examples.

The plasma panel consists of two glass plates spaced by nearly 100 microns. This space is filled with a gaseous mixture containing neon and xenon. When the gas is electrically excited, electrons orbiting the nucleus are extracted and freed. The term "plasma" refers to such a gas in an excited state. The electrodes are silk-screen printed on each of the two plates of the panel, which are line electrodes for one plate and column electrodes for the other plate. The number of line and column electrodes corresponds to the sharpness of the panel. During fabrication, the barrier system is placed in a location that physically makes it possible to define the cell range of the panel and limit the diffusion phenomenon from one color to another. Each intersection of the column electrode and the line electrode will correspond to a video cell containing a large amount of gas. The cell will be referred to as a red, green, blue cell depending on the luminescent deposits that it is covering. Since a video pixel consists of three cells (one red, one green and one blue), there are therefore three times as many column electrodes in a line as pixels. On the other hand, the number of line electrodes is equal to the number of lines in the panel. If such a matrix structure is provided, a potential difference only needs to be applied at the intersection of the line electrode and the column electrode in order to obtain a gas in a plasma state point-wise by exciting a particular cell. The UV generated when the gas is excited will give a cell that emits red, green or blue by impacting the red, green or blue phosphor.

The lines of the plasma panel are addressed as many times as the sub-scans defined in the panel according to the gray level information to be transmitted to the pixels, as described later. The pixel is selected by transmitting a voltage, called a write pulse, through the supply circuit to the entire line corresponding to the selected pixel, and at the same time information corresponding to the gray level value of the selected pixel is parallel to all the electrodes of the column in which the pixel is located. Is sent. Every column, that is, each column, is supplied simultaneously with a value corresponding to the selected pixel of that column.

Therefore, there are each bit of gray level information associated with the time information item corresponding to the bit emission time or, more generally, the time between two writings: a bit with a weight (4) while having a value (1) Will correspond to the pixel being emitted for a duration four times greater than the emission corresponding to the bit with weight 1. The holding time is limited by the time for separating the write signal from the erase cue, and in particular corresponds to the holding voltage which makes it possible to hold the excitation of the cell after addressing the cell. For gray levels coded with n bits (including gray levels for each component (RGB)), the panel will be scanned n times to retranscribe that level, and the duration of each of the sub-scans will be It is proportional to the bit indicated. Through integration, the eye converts the "global" duration corresponding to n bits into emission level values. Therefore, successive scanning of each binary word bit is performed by applying a duration proportional to the weight. The addressing time of a pixel for one bit is the same regardless of the weight of that bit, and what changes is the light emission holding time for that bit.

Thus, inclusive, the cell has only two states, the excited or non-excited state. Therefore, unlike CRT, it is not possible to perform analog modulation of the emitted light level. In order to account for several gray levels, it is necessary to perform a temporal modulation of the emission duration of the cell within the frame period (denoted T). The frame period is generally divided into as many sub-periods (sub-scans) as there are bits (number of bits denoted n) present for coding the video. It should be possible to reconstruct all gray levels between 0 and 255 through the combination based on the n sub-periods. The observer's eye will recreate the desired gray level by integrating the n sub-periods over one frame period.

The panel consists of Nl lines and Nc columns supplied by an Nl line supply circuit and an Nc heat supply circuit. Generation of gray levels by temporal modulation requires that the panel be addressed n times for each pixel of each line. The matrix aspect of the panel will be able to address all the pixels of the same line at the same time by sending an electrical pulse with level Vccy to the line supply circuit. The signal sent to the column is referred to as the column control word and relates to the video signal to be displayed, which is, for example, transcoding which depends on the number of bits used. Video information corresponding to the bits of the column control word addressed at that moment (corresponding to the sub-scan) will be provided to each of the columns, indicating the "binary" amplitude {0 or Vccx (which indicates the state of the coded bits). It will be specified by an electric pulse having. The conjugation of two voltages (Vccx and Vccy) at each electrode junction will or will not induce excitation of the cell. Next, such an excited state will be maintained over a duration proportional to the weight of the sub-scan being performed. The operation will be repeated for every address n and every line Nl addressed. Therefore, it is necessary to address n x N1 lines over the duration of the frame, and thus will provide a basic relationship such as T≥nN _l .t _ad , where t _ad is the time required to address the line.

The sequencing algorithm makes it possible to address every line n times, while following each weighting of the sub-scans performed during each addressing.

Reference is made to FIG. 2 to provide a better explanation of the contouring phenomenon.

In the figure, the abscissa axis represents time and is divided into frame periods having a duration T. Each frame period is a video level (1, 2, 4,) to be displayed on a plasma screen having eight sub-scans by addressing the video quantized to 8 bits so that the duration is proportional to the weight of several sub-scans. 8 ..., 128 are divided into sub-periods with time which makes it possible to define.

The ordinate axis represents the unlit or lit state of the cell as a function of time, for a zero or one level of addressing bits, ie for a given coding level, during the corresponding frame period.

Curve 5 corresponds to the coding of value 128, curve 6 corresponds to the coding of value 127, and curve 7 corresponds to the coding of value 128 and the second frame during the first frame. Corresponds to the coding of the value 127 during the next two frames and vice versa.

The gray level temporal modulation principle involves the temporal distribution of n sub-scans that retranscribe the video over 20 ms of frame. If addressing for eight sub-scans (n = 8) is used, transitions 127/128 and 128/127 involve a transition over every bit. Since eight sub-scans are distributed over 20 ms of a frame, the eye is black, which is part (b) of curve 7 corresponding to zero level for the duration of two consecutive frames, by integrating the video asynchronously. A black area and the appearance of a white area, which is part (a) of the curve 7 corresponding to one level, occur for the duration of two consecutive frames.

The contouring phenomenon appears especially in the moving region where there is a strong transition (contour of the object), more particularly a switchover, at high weighted levels in the coding of the video. In the case of a color screen, this is indicated by the appearance of "wrong color" on the panel due to the incorrect interpretation of the three R G Bs in the contour area. Therefore, the phenomenon is related to the system for temporal modulation of the video level and to the fact that the eye produces the appearance of incorrect contours in its role as an integrator.

The solution to this problem is to define more sub-scans to achieve a better temporal distribution of the information by coding gray levels to be transmitted in more bits than theoretically needed (8 bits to code 256 levels). It is. This is because by increasing the number of sub-scans, each weight of the sub-scans is reduced and problems during their transition are limited. At present, it is possible to perform ten sub-scans (n = 10) within 20 ms, given the characteristics of the panel (the number N1 of lines) and the time t _ad required to address the lines. Gray level transcoding would be 1 2 4 8 16 32 32 32 64 64, for example.

Therefore, the highest weight may be 64 instead of 128.

However, the solution is applied to the quality impairment of the image, which results in limited resolution.

In order to provide a gray level on the plasma panel, it is necessary to perform temporal modulation of the level by performing n consecutive sub-scans during one frame. The sequencing algorithm for addressing is nested, leading to n sub-scan runs of the panel. However, for the purpose of simplifying the algorithm and apparatus for implementing addressing, line 1 + 1 is always addressed immediately after line 1 for a given sub-scan.

The sequencing algorithm according to the prior art is shown in FIG. 3 and will be described later to assist in understanding the present invention by explaining the differences from the prior art.

The sequencing algorithm is known under the title of simultaneous addressing scanning, or SAS (Simultaneous Addressing Scanning). It makes it possible to address every line n times (corresponding to the number n of bits) while following a duration corresponding to the weight of the bit associated with that addressing during each addressing. Each line is addressed for each sub-scan in a defined order as shown in FIG. 3, and for a system with four sub-scans.

The horizontal axis represents time t and the vertical axis represents line number. On the time axis the periods corresponding to the various sub-scans SB0 to SB3 for bits 0 to 3 of the column control word defining the luminance value to be displayed are indicated. The duration of the display, i.e., the sustain duration after recording, actually depends on the weight of the bits of the control word. The duration is an oblique line consisting of two solid lines located on the sides of each of the labels SB0 to SB3, such as for each of the bits 0 to 3, the holding duration indicated as 8 for the sub-scan SB3 for example. Each is marked with. The shaded regions 9 and 11 correspond to the scanning of the previous frame and the subsequent frame, and the middle region 10 corresponds to the scanning of the current frame.

Thus, for a given sub-scan, it becomes apparent that the lines are addressed in ascending order. On the other hand, there are nestings of several sub-scans, which for example are the continuous addressing of the lines from the top of the panel for sub-scan SB1 and the lower panel for the sub-scan SB2 after that moment. It implies that there is a continuous addressing of the line from. In fact, four consecutive lines are subsequently addressed within an addressing period that sends four write pulses before the sustain period.

Thus, if, for example, considering the vertical band 12 corresponding to the short instant dt, the intersection with the oblique line is written on the sub-scans SB3, SB2, SB1 and SB0 of the same frame (in this example). Intersections that represent the starting point in succession and which are indicated on the ordinate axis are, for example, line numbers l ₃ , l ₃ +1, l ₃ +2, l ₃ +3 which are 100 and subsequent lines 101, 102 and 103 for SB3. ), And SB2 for l ₂ , l ₂ +1, l ₂ +2, l ₂ +3, and so on. Addressing of the 4 x 4 lines is performed during the time interval dt. After that moment, we will write lines 104, 105, 106, and 107 for SB3 and will continue to do so.

The new addressing method, which is the subject of this patent application, makes it possible to carry out the writing of lines l and l + 1 at several instants (not consecutive). In practice, the method involves nesting two addressing algorithms, one for even lines and one for odd lines. Overall, everything occurs as if there are currently 2 * n sub-scan algorithms for the N 1/2 line, rather than n sub-scan algorithms for the N1 line. During the addressing period, instead of addressing four consecutive lines (l, l + 1, l + 2, l + 3), now every two lines according to line parity, i.e. (l, l + 2, l +4, l + 6) or (l + 1, l + 3, l + 5, l + 7). This change in addressing is mainly related to the generation of sequencing for the addressing of the various sub-scans.

4 shows how the two addressing algorithms are nested in time. Everything happens as if there are eight sub-scans, each of which applies only to one line parity (even or odd).

The diagonal lines made of solid lines correspond to the sub-scans SB0 to SB3, and the diagonal lines made of dotted lines correspond to the sub-scans SB'0 to SB'3.

As an example, at the instant t, the line addressed for the sub-scan SB3 is an even line l ₃ {actually four consecutive even lines l ₃ , l ₃ +2, l ₃ +4, l ₃ +6)}, and the addressed line for sub-scan SB'2 is an odd line l ' ₂ (actually four odd lines l' ₂ , l ' ₂ +2, l'). ₂ +4, l ' ₂ +6)} and continue to follow the same procedure for other sub-scans at that moment (t).

If the even line l ₃ is written at the instant t, it is observed that the subsequent odd lines l ' ₃ = l ₃ +1 are written at another instant t'.

Therefore, the system for separating the addressing of lines 1 and 1 + 1 implies that the addressing moments of the lines are different. As a result, when line l is addressed, line l + 1 takes a sustain phase. At that moment it is practically possible to turn off the lines l and l + 1 and record the same video information on both lines as described later. In the same way, it is possible to write information only to line l, in which case the holding state for line l + 1 will not be disturbed.

Nesting the sub-scan SB 'to the sub-scan SB may be completely arbitrary, and the two types of sub-scan moments (SB type sub-scan for even lines and SB for odd lines) There is no need for any correlation between types of sub-scans. In the same way, the sustain duration can be completely uncorrelated and can only depend on the weight of the bits of the column control word to be assigned to each type of sub-scan. The weight of the column control word may be selected differently for the sub-scan SB and the sub-scan SB '.

5 and 6 show timing diagrams for two consecutive lines (l and l + 1) and write instants (W) for these lines.

A label of type SB1 means that it belongs to sub-scan 1 (bit n = 1) for a sub-scan of SB type.

T1 represents the sustain duration of the corresponding sub-scan SB1 (bit n = 1).

The arrow shown on the line W1 corresponds to the instant of writing to the line 1.

Line 1 + 1 is controlled by nested sub-scan SB 'as described above.

A label of type SB'1 means that it belongs to sub-scan 1 (bit n = 1) for a sub-scan of SB 'type.

T'1 represents the sustain duration of the corresponding sub-scan SB'1 (bit n = 1).

The arrow appearing on line W1 + 1 corresponds to the instant of writing to line 1 + 1.

The view of FIG. 5 should be contrasted with the view of FIG. 4. In the case of FIG. 5,

In line (l):

Sub-scan (2) (SB2) lasts for T2,

Sub-scan (3) (SB3) lasts for T3,

At line (l + 1):

Sub-scan (1) (SB'1) lasts for T'1,

Sub-scan (2) (SB'2) lasts for T'2,

Sub-scan (3) (SB'3) lasts for T'3.

The order of writing is specific for a single line, and the duration of the sub-scans is independent from one line to another.

Referring to FIG. 4 and taking into account an instant t as an example, it is permissible to start sub-scan SB3 following sub-scan SB2 for line l ₃ . In the next line (l ₃ +1), the instant t overlaps the sub-scan SB2 and the sub-scan SB3 as shown in FIG. 5 and the sub-scan SB'2. Is in the middle.

FIG. 6 no longer refers to FIG. 4 and provides a configuration of the present invention using nested scanning in a general manner.

The first timing diagram corresponds to line 1 and shows four consecutive sub-scans Sb1 to Sb4 with sustain durations t1 to t4.

The second timing diagram corresponds to line 1 + 1 and shows four consecutive sub-scans Sb'1 to Sb'4 with sustain durations t'1 to t'4.

In line (l):

Sub-scan (1) (Sb1) lasts for t1,

Sub-scan (Sb2) lasts for t2,

Sub-scan (3) (Sb3) lasts for t3,

Sub-scan 4 (Sb4) lasts for t4,

At line (l + 1):

Sub-scan (Sb'1) lasts for t'1,

Sub-scan (Sb'2) lasts for t'2,

Sub-scan (Sb'3) lasts for t'3,

Sub-scan 4 (Sb'4) lasts for t'4.

In order to go from the first case (FIG. 5) to the second case (FIG. 6), the three write signals specified in the first case (for each line) become common (two lines l and l + 1) is enough. The remaining recording signals are enclosed in circles in FIG. 6 and designated by reference numeral 13.

Thus, by adding the remaining first and second write signals to line 1 (the erase signal for the previous sub-scan always precedes), the sustain duration T2 is divided into two periods t1 and t2 and , The sustain duration T3 is divided into two periods t3 and t4.

For the next line l + 1, the addition of the write signal makes it possible to divide the sustain duration T'2 into two periods t'2 and t'3.

By based on nested SB and SB 'type sub-scans, it is possible to reduce to both sub-scans common to lines 1 and 1 + 1 through both duration and video content (0 or 1). Thus, it is possible to perform line copyover. The line copy will be dubbed "partially" as long as it is performed upon request. In particular, for example, the operation performed in three writes can be reduced to zero (this is the first case) and can be reduced to one write or two writes.

The representation of the partial copy on request is used due to the introduction of the concept of variable parameters that can be defined as a function of the video content.

The great advantage of the method lies in the fact that it is easy to go from 16 sub-scan modes to 13 sub-scan modes (see example provided below) from one frame to another without any transition period. Therefore, such adaptation can be performed in accordance with the content function of the sequence and even according to the content function of the image. A system for measuring vertical resolution can be used to make a decision on the number of sub-scans to be used. The method even makes it possible to go from a pair of lines to another pair of lines, ie from 13 sub-scan mode to 16 sub-scan mode. Decision information may be calculated for each line pair.

Next, a configuration for separating the information into common values and specific values will be specified, which can be combined with the present invention.

The coding of the gray level according to its configuration, specified by the column control word, is performed by taking into account the luminance value of the selected pixel as well as the luminance value of the pixels located in adjacent lines for the same column.

In practice, the column control word is divided into two parts, a first control word corresponding to a value common to two pixels and a second and third control word corresponding to a specific value of the pixel, for a given pixel.

It is desirable to obtain the following coding:

n1 + n2 + n3 = 2 × (number of sub-scans per line)

A value specific for line l coded with n1 bits

A value specific for the line (l + l) coded with n2 bits

A value specific for lines l and l + 1 coded with n3 bits.

If a given number of sub-scans is taken into account, in practice, it relates to the bits for coding two specific values and the common values, to the coding bits for line l, and to the coding bits for line l + 1. The number of sub-scans involved, i.e., n1 + n2 + n3, needs to correspond to the number of sub-scans performed in a conventional manner.

The various parameters n1, n2, n3 are not fixed. It is possible to adjust the relationship between the definition of a specific value and the definition of a column value. The better the particular value is defined, the smaller the loss of resolution associated with coding will be. Conversely, the smaller the specific value is defined, the larger the total number of sub-scans will be. Therefore, there is a compromise to be established between loss of resolution on the one hand and minimization of display defects on the other hand.

The calculation of a specific value is performed as follows:

The specific value for lines l and l + 1 include an item of information about the difference between the lines l and l + 1. This means that if NG1 and NG2 represent gray levels for the pixels of lines l and l + 1, VS1 and VS2 represent their specific values, and VC represent common values,

NG1 = VS1 + VC

This is because relations such as NG2 = VS2 + VC are maintained.

Therefore, VS1-VS2 must be equal to NG1-NG2 (to always have zero coding error). When the difference (denoted D) of NG1 and NG2 is determined, VS1 and VS2 are calculated by adding the term D and the lowest gray level portion α.

If NG1> NG2 VS1 = D + αNG2

VS2 = αNG2

If NG2> NG1 VS1 = αNG1

VS2 = D + αNG1

The value of α is a parameter to be defined in the same manner as n1, n2, n3. The value α is the result of a computational test and thus is determined in part experimentally. The value is selected according to the derived calculation function, for example a value of 3/16 that facilitates calculation by a digital signal processor (DSP).

The common value is calculated by subtracting the starting value from the specific value. If an approximation is used to calculate a particular value, then the common value is obtained through a calculation such as VC = 1/2 × (NG1 + NG2-VS1-VS2).

Therefore, the calculation is summarized in the following steps:

Determining a value D corresponding to the difference between the two values NG1 and NG2 to be coded,

Calculating specific values VS1 and VS2 as a function of D, α and NG1 or NG2,

Calculating a common value VC as a function of NG1, NG2, VS1, VS2.

The important point is to minimize the recording error. In order to minimize the recording error, use will be made for specific coding of specific values. This is coding with an increase of five, ie each code is a multiple of five. The following table shows how certain values and common values are calculated to obtain the values VF1 and VF2 that are as close as possible to the final possible NG1 and NG2. In practice, the errors E1 and E2 are limited to +/- 1.

NG1 NG2 D D which is a multiple of 5 VS1 VS2 VC VF1 VF2 E1 E2 60 65 5 5 10 15 50 60 65 0 0 60 66 6 5 10 15 50 60 65 0 -One 60 67 7 5 10 15 51 61 66 One -One 60 68 8 10 10 20 49 59 69 -One One 60 69 9 10 10 20 49 59 69 -One 0

The difference D between gray values is coded based on the closest multiple of 5 of that value D. The specific values VS1 and VS2 are multiples of 5, and the ratio of the specific value to the global value (parameter α) is selected equal to 3/16. Therefore, the value of VS1 is the value of modulo 5 which is closest to 60 x 3/16.

The particular value containing the item of information about the difference between the two coded pixels is only limited through a limited number of bits. Therefore, the maximum difference that will be possible to code is, in fact, limited to the maximum value that can be coded with a particular value. Thus, this would forbid coding a large difference.

For strong transitions, because the difference that can be coded is limited, one of the specific values will be equal to the maximum value and the other value will be equal to zero. The common value will be determined in such a way as to minimize the error in the final value. In this case, the final error may be greater than one.

The following table provides an example of coding between two pixels, where the difference between the two pixels is greater than the maximum limit of a particular value. The maximum value selected for a particular value is taken equally to be 70.

NG1 NG2 D D is a multiple of five VS1 VS2 VC VF1 VF2 E1 E2 10 100 90 70 0 70 20 20 90 10 -10

An example application implementing a configuration to separate that information into common and specific values is provided below for a system that allows 10 sub-scans:

Parameter definitions:

N1 = 4 (codes 5, 10, 20, 35)

N2 = 4 (codes 5, 10, 20, 35)

N3 = 12 (codes 1, 2, 4, 6, 9, 12, 15, 19, 23, 27, 31, 36)

Α = 3/16

This actually copies the gray level into 16 sub-scans, i.e. 12 sub-scans common to both lines (and therefore equivalent to 6 conventional sub-scans) and 4 specific sub-scans. In that case, the gain would be six sub-scans with a recording error less than or equal to 1 (for the difference between lines less than or equal to 70).

7 shows such addressing with 16 sub-scans. The sub-scans corresponding to the bits with weights 10, 9, 15, 12, 20 follow each other as a function of time in line 1 and line 1 + 1. The record labeled 14 is common to lines l and l + 1 for values 9, 15 and 12. The record, denoted by reference numeral 15, is specific for lines l and l + 1 and relates to values 10 and 20.

Thus, the limited 16-bit code limits the maximum difference between lines 1 and 1 + 1 to 70 (70 = 5 + 10 + 20 + 35). Above 70, coding operations in 16 bits involves the generation of errors larger than the LSB.

The problem is solved by combining the above-described configuration with a configuration for nesting sub-scans.

The 16-bit code above corresponds to the weight of the bits of the column control word computed from the video information:

1 2 4 5 6 9 10 12 15 19 20 23 27 31 35 36

According to the configuration for separating the information into a common value and a specific value, each video information item is divided into an information item specific to the current line l and an information item common to two adjacent lines l and l + 1. do. Specific information is coded into four bits, each weighting a multiple of five (5, 10, 20, 35). Common information is coded 12 bits.

The sub-scan nesting configuration makes it possible to increase the value of the maximum difference in a direction where errors are no longer ignored, which is particularly useful when the vertical resolution (difference in brightness) is quite large.

It makes it possible to dynamically go from 16 sub-scans (10 sub-scans and 4 separate sub-scans common to both lines) to 13 sub-scans.

First, each order of the various sub-scans is adjusted as follows:

1 2 4 6 5 10 9 15 12 20 19 23 27 31 36 35

The order defines the rank of the bits of the control word that are transmitted and represented by their weights.

The first four sub-scans (1, 2, 4, 6) are always common for two adjacent lines. The sub-scans 5 and 10 and also 20, 35 are always specific to them for lines l and l + 1 (hence there are always two different information items for the sub-scan). .

For the next three sub-scans 9, 15, 12, two cases are possible, where the sub-scans are common for two lines (and then one reverts to 16 sub-scan addressing) Or when the sub-scans are partially specific (13, 14 or 15 sub-scan addressing).

8 illustrates such 13 sub-scan addressing. The sub-scans corresponding to the bits with weights 10, 24, 12, 20 follow each other in line 1. The sub-scans corresponding to the bits with weights 10, 9, 27, 20 follow each other on line 1 + 1. The record, indicated by reference numeral 16, is common to lines l and l + 1 for values 9 and 24. The record, denoted by reference numeral 17, is specific for lines 1 and 1 + 1 and relates to values 10, 20, 12 and 27. In practice, it is a write relative to the common sub-scan 9, but line l will not be erased at the end of the sustain period. If there is no erasure, the recorded information still exists, which implies that the video information with weight 9 in line 1 + 1 has a different weight 24 in line 1. On the other hand, the line 1 + 1 is erased at the end of the period with the weight 9. At that moment, the next video information item will be written to line l + 1 (corresponding to 15 in 16 sub-scan mode). In the same way, line 1 will be erased at the end of the period with weight 15 rather than line 1 + 1. Accordingly, there is a sub-scan in line 1 with a duration 24 (9 + 15) equal to the sub-scan in which video content has a duration 9 of line 1 + 1. Next, the video content of sub-scan 12 is recorded in line 1. In the same way, when writing 12 in line 1, no erase is performed in line 1 + 1. As a result, the sub-scan 15 of line l + 1 actually lasts for 27 (15 + 12). Next, the erasing signal common to the lines l and l + 1 is performed before recording the video information corresponding to the specific weight value 20.

In conclusion, in the 16 sub-scan modes, there were three consecutive common sub-scans with respective weights (9, 15, 12). In the 13 sub-scan mode, there are actually two sub-scans 24 and 12 in line l and two sub-scans 9 and 27 in line l + 1. Only constraint, information item 24 of line l and information item 9 of line l + 1 are common. On the other hand, the weight 12 of line l and the weight 27 of line l + 1 are specific. Therefore, the ratio of certain values is increased relative to the common value, which allows for greater vertical resolution.

In the same way, the sub-scans 19, 23, 27, 31, 36 with 16 sub-scan addressing have three sub-scans 42, 58, 36 and line l + 1 for line l. It can be transformed into three sub-scans 19, 50, 67 for. Only limitedly, the video information of the sub-scan 42 of line 1 is the same as the video information of the sub-scan 19 of line 1 + 1.

Saving of one sub-scan was made for the coding of values 9, 15, and 12, and saving of the other two sub-scans was made for coding of values 19, 23, 27, 31, and 36.

To consider the average number of sub-scans per line by taking into account common sub-scans and specific sub-scans for two lines, let's calculate the number of writes for two consecutive lines:

4 records corresponding to 4 common sub-scans (1, 2, 4, 6),

4 x 2 records corresponding to 4 specific sub-scans (5, 10, 20, 35),

1 record corresponding to 1 common sub-scan (9 + 15 for line l and 9 records for line l + 1) 1 record common to both lines for 9 sub-scan 9 Limited to control},

1 x 2 writes corresponding to 2 specific sub-scans (12 for line l and 15 + 12 for line l + 1),

One record corresponding to one common sub-scan (19 + 23 for line l and 19 for line l + 1) one record common to both lines for sub-scan 19 Limited to control},

1 × 2 writes corresponding to 2 specific sub-scans (27 + 31 for line l, 23 + 27 for line l + 1),

1 x 2 writes corresponding to 2 specific sub-scans (36 for line l, 31 + 36 for line l + 1).

That is, a total of 4 + 8 + 1 + 2 + 1 + 2 + 2 = 20 records are calculated.

The average of 10 records per line is clearly set.

Alternatively, it may be mentioned that the column control word was coded with 16 bits, and, depending on the weight of the bits, the lines were addressed individually or by two lines. Thus, writing two bits where the lines were addressed by two lines The scan time for doing so is halved and the scan time is reduced up to the scan time of a ten word (4 + 12/2) ten words.

According to the sub-scan nesting configuration, the column control word is coded into 13 bits, some of which are common to two consecutive lines.

The column control word has bits with different weights depending on whether the associated line is an even or an odd line.

The weight of a column control word coded in 13 bits (13 sub-scans) is:

For even lines (or odd lines depending on the selection of the lines),

1, 2, 4, 6, 5, 10, 24, 12, 20, 42, 58, 36, 35,

For odd lines (each even line),

1, 2, 4, 6, 5, 10, 9, 27, 20, 19, 50, 67, 35.

The weights of the bits of ranks 7 and 8 have the same sum 36. The weights of the bits of ranks 10, 11, 12 have the same sum 136.

The lines, in one example,

It is addressed by two lines for a weight of 1, 2, 4, 6, 9 or 24, 19 or 42 (according to the relevant column control word).

The lines are addressed separately for the weights 5, 10, 20, 35.

The lines are individually addressed for weights {(15 + 12), (23 + 27), (31 + 36)}.

The lines are individually addressed for weights {12, (27 + 31), 36}.

The scan time for a record that corresponds explicitly to 10 bits is obtained:

4/2 + 2/2 + 4 + 6/2 = 10

Overall, the present invention goes from a maximum difference of 70 for 16 sub-scans to a difference of 176 (255-42-24-13) for 13 sub-scans. Therefore, this makes it possible to significantly increase the transmitted vertical resolution (values of 9/24 and 19/42, since the weights 1, 2, 4, 6 cannot actually be connected individually). }.

The great benefit of this technique is that it is possible to perform switching between 16 sub-scan addressing and 13 sub-scan addressing on demand for a given line pair. As an example it is possible to detect upstream of an image region with a strong vertical transition. Next, all lines in the area will be subjected to 13 sub-scan addressing, and the rest will probably be at 16 sub-scan addressing. The transition corresponding to going from addressing according to FIG. 8 to addressing according to FIG. 7 selects (or lines) during writing of bits with different weights in line l (or l + 1). The selection of (l + 1) is performed in a simple manner by replacing the selection of line l with the immediately following (or preceding) line for simultaneous writing on both lines.

In the same way, it would be advantageous to have a "wrong loop" detector to assess the need or otherwise to survive in 16 sub-scan mode. There is a compromise to be established between limiting the vertical resolution and the level of "wrong contour".

The number of sub-scans is associated with the number of bits, the number of bits having different weights from the column control word corresponding to the line to the next line and the column control word corresponding to that number, and thus used for coding the image. The column control word to be selected can be selected as a function of the image to be processed, and it is further possible that the selection is made for each image. The weight of the associated bit may be selected according to the resolution function of the image.

The problems of quantization and cell priming described above can be reduced as follows:

By using only the configuration for separating the addressing of lines 1 and 1 + 1, it is possible to improve the priming of the excitation in a clearly simple manner. In particular, during conventional addressing, the four cells addressed during the current period are first turned off by the erase pulse. Subsequent writing immediately following will not benefit from the proximity effect of the lighted cells. Only cells that can emit light are cells located directly above or below a packet of four lines.

In the case of the present invention where lines l and l + 1 are addressed at different instants, line l + 1 can benefit from the possible excitation states of lines l and l + 2, and that line l And l + 2) are not turned off immediately before. In fact, it is possible to get all the sub-scans to benefit from every line of the system.

To facilitate prime of all sub-scans, it is sufficient to have the writing moments on even and odd lines that are systematically different from each other. A simple way to achieve that is to offset the two addressing systems by a certain amount of time while keeping the same code on both lines in that case. As an example it is possible to use a dual addressing system, one offset by the equivalent of 1/2 LSB relative to the other.

In the example of FIG. 8, which is a configuration with thirteen sub-scans, certain sub-scans benefit from the accelerated priming.

For low level quantization, if considering two separate addressing for odd lines and even lines, it is possible to perform a common write on two adjacent lines at a given moment, as mentioned above. This results in stopping the sub-scan holding state of the line 1 and recording the video information of the line 1 + 1 to the lines 1 and 1 + 1. The duration of the initial sub-scan of line 1 is reduced in this case. Applying the configuration to the sub-scans corresponding to the lowest weight (duration corresponding to the LSB) results in smaller quantization intervals than the LSB. The phase shift between the two addressing may be chosen to be equal to 1/2 LSB. If a common addressing scheme is applied to two adjacent lines, then a sub-scan with a weight (1/2 LSB) is defined. This obtains a quantization level that can be used specifically for low levels. It is also possible to define an addressing system that makes it possible to further increase its quantization by deriving a weight that is one quarter of the LSB.

An exemplary embodiment of an apparatus implementing the scanning method is described below. A simple chart of the control circuit of the plasma panel 18 is shown in FIG. 9.

The digital video information arrives at the input E of the device, which is also the input of the video processing circuit 19 and the input of the selection circuit 20 based on the microprocessor. The video processing circuitry is also referred to as a scan generator or circuit 24 (hereinafter also referred to as a scanning circuit or control circuit) for controlling the corresponding memory 21, the selection circuit 20, the input end of the video memory 22, and the line supply circuit. ). The video memory transmits the stored information to the input of the circuit 23, which heat supply circuits are grouped together.

The scan generator 24 transmits synchronization information to the video memory 22 and controls the circuit 25 in which the line supply circuits are grouped together.

Thus, the video information coded in 8 bits and received on input E is sent to selection circuit 20 which stores the video data for the complete image. The circuit analyzes the content of the video and calculates the number of times a luminance difference between lines l and l + 1 that is greater than a preset threshold exists in the image.

If the number is greater than the predetermined threshold, the scan is performed using the sub-scan nesting configuration, ie based on 13 sub-scan addressing. In the opposite case, 16 sub-scans are performed. Information about that type of scan is sent to processing circuitry 19, which executes the coding of the video information accordingly. The processing circuit transmits the information to the scanning circuit 24 so that the scanning circuit 24 executes scanning of the screen as a function of its coding.

The processing circuit 19 exchanges the video data with the corresponding memory or table 21, which will provide the word as data, as a function of the value of the video word sent according to the address. The word corresponds to a 13 or 16 bit code whose weight will be predefined. The transcoding from the correspondence table 21 is defined as a function of the addressing mode used.

When the 13 addressing sub-scan mode is selected, a 13-bit coded word corresponds to the following two types of coding, separated by the weight of the bits of the coded word:

A first coding type providing a first coding word corresponding to an even line of the plasma panel,

A second coding type providing a second coding word corresponding to an odd line of the plasma panel.

The word is then sent to video memory 22, which stores the word to provide the column supply circuit with consecutive bits of the column control word in synchronism with the line scan.

The scan generator 24 performs line scanning of the screen via the line supply circuit 25 over the duration of one frame. The circuit 25 provides an addressing voltage and also provides a sustain voltage for a duration corresponding to a sub-scan for the weight of the bits sent to the column for that addressing.

The scan generator 24 performs sub-scans according to the command function received from the processing circuit.

The types of scans implemented are:

Scan of the selected lines by two lines (simultaneous selection of lines 2l and 2l + 1),

-Scan of each continuous line.

The transition from the 13 sub-scan mode to the 16 sub-scan mode does not select only the line 2l or only the line 2l + 1 when writing the bit corresponding to the common value VC. And 2l + 1) very simply.

It should be noted that the selection circuit 20 can be reliably placed upstream of the device and in particular the processing circuit in order to avoid any delay in the coding of the video word.

Of course, the above description assumes line selection of the plasma panel for transmitting video information on the column input of the display, but other types of addressing may be used as a function of the lines and columns without departing from the method from the scope of the present invention. Can be envisioned by changing.

Certainly, the present invention is not limited by the number of bits to quantize the digital video signal to be displayed, but also by the number of sub-scans.

Matrix addressing using temporal type of modulation for display of luminance or gray level corresponding to each of the three components R G B can equally be applied to any type of screen or device. The cells of a matrix array or device having a line input and a column input, where the term cell is broadly considered to be elements at the intersection of lines and columns, may be cells of a plasma panel or otherwise micromirrors in a micromirror circuit Can be. Instead of emitting light directly, the micromirror reflects the received light in a pointwise manner (cell corresponding to the micromirror) when they are selected. Next, their addressing for selection is the same as the addressing of the cells of the plasma panel as described in this application.

Claims (13)

A method for addressing cells arranged in a matrix array, Each cell is located at the intersection of a line and a column, and the array has a line input stage and a column input stage for displaying a gray level defined by a video word that forms a digital video signal and defines an image, each column input stage Receive a control word for the column corresponding to the video word for the line addressed for the column, wherein the word transmitted continuously consists of n bits, each sequence corresponds to a sub-scan, and each bit A method for addressing cells, with or without triggering, during a time proportional to the weight of the bit in the word, according to its state, i.e., light emission of a cell in the addressed line and column receiving the control word. To The other coding of the column control word is performed depending on whether it is related to the even or odd lines of the word, the difference being that at least m consecutive specific rank bits, between 2 and n, from one control word to another control word. With a different weight, and the sum of the weighted weights of the bits lasts equally from one control word to another, thereby obtaining substantially different write instants from one line to the next. Method for addressing arranged cells. 2. The method of claim 1, wherein the write is simultaneous to two consecutive lines for at least the first of m consecutive bits of the control word for one of the two lines. Way. 3. The method of claim 1, wherein at least two consecutive lines are simultaneously selected for at least one of the bits of a particular rank having the same weight from one control word to the other. A method for addressing cells arranged in a matrix array. 4. A method according to any one of the preceding claims, wherein at least one of said bits of a particular rank codes a partial luminance value common to two consecutive lines, having the same weight from one control word to another control word. Wherein the write is simultaneous to the lines with respect to the bit of the control word on one of the two lines. The method of claim 1, wherein the method is implemented for a limited number of the matrix array lines, the lines corresponding to the picture zone defined by the video signal having a strong vertical transition, and the other zone is the thermal control. A method for addressing cells arranged in a matrix array, characterized in that all words use a sub-scan corresponding to the same weighting addressing method from one line to another. The method of claim 1, wherein the method is implemented for an image having a strong vertical transition, and another image uses an addressing method in which all of the column control words have the same weight from one line to another line. A method for addressing cells arranged in a matrix array. 2. The method of claim 1, wherein from the first addressing method comprising n sub-scans, a greater number of sub-scans and wherein the column control word has the same weight from one line to another line The switch to the second addressing method with bits comprises selecting the line (l) along with writing the bits with different weights on the line (l) in the first method, in the second method. 1) and by replacing the line immediately preceding or immediately following the line (l) for simultaneous writing on both lines, either immediately preceding or following the line. A method for addressing cells arranged in a matrix array. 2. The method of claim 1, wherein the value of m or the value of the weight corresponding to the m bit depends on the vertical resolution of the image. 9. A method according to any one of the preceding claims, wherein said cells are cells of a plasma panel and said selection causes light emission of said cells. 10. The method of any one of the preceding claims, wherein the cells are micromirrors of a micromirror circuit. Video processing circuitry 19 for processing the received video data, A corresponding memory 21 for transcoding the data; A video memory (22) for storing said transcoded data, said video memory (22) being linked to a column supply circuit (23) for controlling column addressing of a plasma panel based on a column control word; An apparatus for implementing the method according to claim 1 comprising a control circuit 24 for the line supply circuit 25, which is linked to the video processing circuit to select a line. The video processing and transcoding circuit performs different coding of the column control word depending on whether the word is related to even or odd lines, the difference being at least m consecutive between 2 and n of the bits to be transmitted. A particular rank with an average bit (m is between 2 and n) has different weights from one control word to another control word, and the weighted sum of the bits survives equally from one control word to another control word, Device for obtaining substantially different recording instants from one line to the next. 12. The circuit of claim 11, wherein the circuit for controlling the line supply circuit selects two consecutive lines simultaneously while transmitting by the column supply circuit the first bit of the consecutive bits of a control word for one of the two lines. Device, characterized in that. 12. The apparatus of claim 11, wherein the apparatus corresponds to addressing according to n sub-scans or addressing corresponding to a greater number of sub-scans, according to a luminance variation function from one line to another line in the image or portion of the image. And selection circuitry (20) for receiving said video data to select a coding of said column control word.