JP3425083B2 - Image display device and image evaluation device - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image display device using a display panel such as a plasma display panel for performing multi-gradation display by dividing one TV field of an image into a plurality of subfields and displaying the divided subfields. This is a technology related to image quality improvement. The present invention also relates to an image evaluation device that evaluates a display image in such a moving image display device.

[0002]

2. Description of the Related Art Plasma display panels (plasma d)
An image display device using a display panel that emits light in a binary manner represented by an isplay panel, hereinafter simply referred to as "PDP", is, for example, an address display period separated sub-field met.
The gradation display is realized by a display method called hod). In this system, 1 TV field is converted to 1 of PDP screen.
The image display is performed by time-division into several subfields each including an address period in which lighting / non-lighting data is written for each line, and a discharge sustaining period in which a predetermined pixel emits light all at once.

Conventionally, in the case of performing multi-gradation display by dividing one TV field of an image into a plurality of sub-field images for display in this way, so-called pseudo contour-like gradation disturbance occurs in moving image display. Is known to do.

The generation of the pseudo contour when the moving image is displayed will be described with reference to FIGS. 35 and 36. FIG. 35 shows 127 and 128.
4 that have adjacent gradation levels between adjacent pixels
The image pattern PA1 including one pixel is the PDP 300.
2 shows a state in which the screen is translated by 2 pixels in one TV field. Further, in FIG. 36, the horizontal axis represents the relative position of each pixel on the screen, and the vertical axis represents only the time corresponding to one TV field for convenience. Further, it shows a state where the observer follows the movement of the image pattern PA1 in parallel. Here, 8-bit gradation, that is, 256 gradations, is converted into lighting / non-lighting 8-bit data of eight subfields, and a corresponding gradation display is performed based on the 8-bit data. ，
A case where one TV field is time-divided into subfields 1 to 8 in this order (ascending order) by weighting 2, 4, 8, 16, 32, 64, 128 will be described. In order to display the gradation level 127, the gradation level 128 is displayed by turning on the subfield 1 to the subfield 7 (hatched portion in the figure) and turning off the subfield 8. In order to display, the sub-fields 1 to 7 are turned off and the sub-field 8 is turned on (hatched portion in the figure) to display the gradation level.

When a still image is displayed, the average brightness of one TV field of the observed image is represented by the time integration of lighting between AA 'in FIG. 36, and gradation display is performed correctly. On the other hand, when a moving image is displayed, depending on the direction of movement of the line of sight, the retina may be displayed on the retina between BB ′ and C− in FIG.
The time integration of lighting between C'is observed. The combined value of each bit (subfield) between BB 'is about 0, and the total of each bit (subfield) between CC' is about 255. In this way, when it is observed that the image patterns adjacent to each other are similar in gradation level such as the gradation level 127 and the gradation level 128,
In the level changing portion, the gradation level observed as shown in FIG. 36 is significantly disturbed by the movement of the image.

That is, since the halftone is expressed by the integration of the luminance of each subfield in the time direction, when the line of sight moves in a moving image or the like, an image at a position different from the original pixel position is displayed as time passes. Since the brightness weights of the respective bits are integrated, the halftone display is greatly disturbed. It should be noted that this disturbance of the halftone is recognized as a false contour appears in the image, and is therefore generally called "moving image pseudo contour". Note that the mechanism of generating the pseudo contour in such a moving image display is described in Reference 199.
Published on May 1, 7th, All about plasma display, 1
This is explained in detail on pages 65 to 177 (published by Kogyo Kenkyukai, published by Hiraki Uchiike and Shigeru Mikoshiba).

[0007]

In order to eliminate the false contour of a moving image, in the conventional image display device, the subfield 7 and the subfield 8 corresponding to a plurality of high-order bits.
An attempt has been made to reduce the disturbance of halftone display in moving image display by dividing the luminance weights of the above and further disposing them in the first half and the second half of the field. FIG. 37 shows a subfield structure in the moving image false contour reducing method according to the conventional method, which is intended to display an 8-bit gradation level, that is, 256 gradation levels by using 10 subfields. The luminance weighting of each subfield is 48, 48,
It is 1, 2, 4, 8, 16, 32, 48, 48. That is, the luminance weights 64 and 128 of the upper 2 bit subfield 7 and the subfield 8 in the above eight subfields are divided into four luminance weights ((64 + 12
8) = 192 = 48 × 4) and disperse these in the first and second halves of one TV field, and this is a technique for reducing the weight of upper bits to suppress the occurrence of intermediate tone disturbance as much as possible. . According to this technique, at the boundary between the gradation levels 127 and 128, almost no disturbance in gradation is observed, and the generation of a moving image pseudo contour is suppressed at that portion. For example, the gradation shown in FIG. Level 63
And 64, a subfield having a large luminance weight (here, subfield 9) is first lit, and a subfield having a small luminance weight being lit (here, subfields 3, 4, 5, 6, and 8). ) Is not illuminated, the distribution of illumination / non-illumination in the subfield changes greatly, so gradation disturbance is still observed at the boundary portion. That is, the gradation level observed in the direction of the dotted arrow Ya is about 79, whereas the gradation level of the dotted arrow Yb is
The gradation level observed at is about 32. Therefore, when a moving image with such a gradation is displayed, the generation of the moving image pseudo contour cannot be suppressed.

Further, in the above-described moving image pseudo contour evaluation method, luminance weights of all subfields existing on the dotted arrow Ya or Yb as shown in FIG. 37 are added to obtain an observed moving image pseudo contour. Because of the calculation, even if the motion of the image changes slightly from the direction indicated by Ya or Yb on the dotted arrow to be evaluated, there is a subfield that deviates from this dotted line, or conversely, it newly enters on the dotted line. There are subfields that come up. This state is shown by the dotted arrow Yc or Yd in FIG. 37, and means that the calculated amount of the moving image pseudo contour is significantly different even if the difference in the movement of the image is slight.
In this way, in the conventional evaluation method, the presence or absence of the addition of the luminance weights of the subfields is calculated by selecting whether it exists on the dotted line or deviates from the dotted line. The image of the pseudo contour evaluation result may be greatly different, and the image actually reflected by the observer cannot be obtained, so that the accurate evaluation cannot be performed.

Further, the assumed movement of the image is only one of the horizontal movement and the vertical movement, and there is a problem that it is difficult to evaluate when the horizontal movement and the vertical movement are taken into consideration at the same time, that is, when the movement is performed in an oblique direction. It was

Next, a CRT display device is now widely used as a television image display device. The CRT display device has a track record as a display device, is low in manufacturing cost, and is highly evaluated for other display performances such as brightness and contrast. However, the size and weight of the entire display device are large, and improvement has been desired as a thin display such as a wall-mounted television. On the other hand, as thin and lightweight display devices, the performance of PDPs and liquid crystal displays has been improved, and therefore display devices using these display devices are gradually receiving attention. Liquid crystal displays are currently considered to be suitable for relatively small-sized display devices, and are widely used particularly as display devices for notebook computers. However, there are problems in that it is still difficult to increase the screen size, and the display response characteristics when a moving image is displayed are insufficient to cause an afterimage. On the other hand, the plasma display is expected to be a wall-mounted television in the future because it is determined that the size of the plasma display is relatively promising.

In a normal CRT display device, when one electron beam irradiates a predetermined pixel, not only the pixel but also peripheral pixels emit light to a considerable extent at the same time, so that the image display information is diffused. However, as a result, the spatial frequency characteristic is deteriorated. On the other hand, in a matrix type display device such as a PDP or a liquid crystal display, the image display information at each pixel and the image display information at an adjacent pixel are independent because each pixel has an individual electrode. Because of its high quality, the image is sharp and has a high reputation such as a clear display. However, as described above, the liquid crystal display device has a drawback that the display response characteristic is insufficient and an afterimage occurs when a moving image is displayed. PDPs are expected to be capable of comprehensive high-quality display because they do not have a delay in response characteristics as in liquid crystal display devices.

By the way, in the conventional image display device using the PDP, the display device is configured using the same signal source and signal processing as in the case of using the conventional CRT display device, except for the PDP portion. Among the noises included in the input video signal, the noises of the two-dimensional high frequency components are not particularly noticeable in the conventional CRT display device, but there is a new problem that the image is conspicuous especially in the still image portion where the image is fine. It was happening.

Therefore, a first object of the present invention is to provide an image display device capable of reducing the occurrence of moving image pseudo contours more than ever before.

A second object of the present invention is to provide an image evaluation apparatus capable of evaluation that reflects an actually visible image regardless of the moving direction of the image.

A third object of the present invention is to provide an image display device capable of displaying an excellent image which is not significantly affected by the noise component of the input video signal.

[0016]

In order to achieve the first object, the brightness weights W1, W2 ,. ． ． , WN,
Then, the signal levels are 0, W1, W2 ,. ． ． , WN can be expressed in any combination, and a predetermined signal level is selected according to the amount of motion (the amount of motion refers to the temporal variation of the input video signals of a plurality of frames), and the selected signal level is selected as the display signal. To do. This can be said to be an extremely effective technique for eliminating the false contour of a moving image with a configuration other than devising the weighting of the subfields and the arrangement thereof. Of course, if the subfield brightness weighting and its arrangement are devised at the same time, the effect of eliminating the pseudo contour becomes more remarkable.

Here, if the difference between the signal level of the input signal and the signal level of the display signal represented by the limitation is distributed to the peripheral pixels, the error between the level of the input signal and the level of the display signal can be almost cancelled.

Further, the following image display device is provided. In other words, one TV field is composed of N subfields each having a luminance weight arranged in time order, and the initialization for the subfield is (N-1).
Do the following times. Then, a lighting method in which a subfield that emits light in proportion to the value of the input video signal extends forward or backward in the time direction within a predetermined range of the input video signal becomes possible, which is the first object of eliminating the pseudo contour. Is achieved.

Next, in order to achieve the second object, the following image evaluation device is provided.

That is, a reference point is set at a predetermined pixel on the display screen, a path passing through the reference point moving in a unit time on the screen is assumed, and light is emitted from a pixel near the path in the assumed unit time. This device is a device for making an observed image at the reference point by performing a predetermined calculation on the emitted light amount and then integrating the same.

According to this, not only one pixel on the path through which the line of sight passes, but also the light emission from a plurality of pixels in the vicinity of the path through which the line of sight passes is taken into consideration. Therefore, even if the expected movement of the image changes slightly, the image of the evaluation result changes greatly.
The instability is eliminated, and the image actually viewed by the observer is reflected, enabling stable image evaluation.

Further, in order to achieve the third object, there is provided an image display device which suppresses a time response of a high frequency component of a spatial frequency component of an input signal to obtain a display signal.

[0023]

DESCRIPTION OF THE PREFERRED EMBODIMENTS [First Embodiment] FIG. 1 is a block diagram showing a configuration of an image display device according to the present embodiment.

As shown in FIG. 1, the image display device according to the present embodiment has a filter section 1, a γ (gamma) inverse correction section 2,
An AD conversion unit 3, an error diffusion unit 4, a motion amount calculation unit 5,
First encoding unit 6, second encoding unit 7, and display control unit 8
And PDP9.

The PDP 9 is a display device in which electrodes are arranged in a matrix and which has, for example, (640 pixels / line) × 480 pixels, and which performs binary light emission such as ON or OFF. Then, gradation is expressed by a total of a predetermined number (for example, 10) of subfields having a predetermined number of times of light emission as a luminance weight, and halftone display is performed. It should be noted that in the present embodiment, a PDP that displays in a single color will be described for the sake of simplicity.
The same can be applied to each color even in a PDP in which pixels are formed by three colors (red), G (green), and B (blue) to perform color display.

The filter unit 1 will be described in detail later, but
This is a circuit for removing high frequency components of spatial frequency.

The γ inverse correction unit 2 has a γ (normally γ = 2.2) characteristic with respect to the original video signal on the assumption that the analog video signal used here is displayed on a CRT. For correcting the display signal and the original input signal to have a linear (γ = 1) input / output relationship.

The AD converter 3 is a circuit for converting an analog video signal into a 12-bit video signal here.

FIG. 2 is a block diagram showing the configuration of the second encoding unit 7.

As shown in this figure, the second encoding unit 7 includes a subfield conversion unit 71 and a write address control unit 72.
And frame memories 73a and 73b.

The write address control section 72 generates an address designation signal for designating a frame memory write address based on the horizontal synchronizing signal and the vertical synchronizing signal separated from the video signal.

The sub-field converter 71 is a circuit for converting the digital video signal corresponding to each pixel into 10-bit field information having a predetermined weighting. In addition, the second encoding unit 7
An 8-bit digital video signal (b) obtained by encoding the signal (a ′) in which the lower 4 bits are truncated by the encoding unit 6 is input.

The field information is a set of 1-bit subfield information indicating which time zone in one TV field, that is, which subfield is turned on or off. Here, by referring to the subfield conversion table 710 stored in the subfield conversion unit 71 according to the gradation level of the input digital video signal, the 8-bit video signal corresponding to each pixel has a predetermined number of subfields. It is divided into fields. The division process for each pixel is performed in synchronization with a pixel clock generated by a PLL circuit (not shown). The field information corresponding to each pixel generated in this manner has a physical address specified by an address specification signal from the write address control unit 72 and is stored in the frame memories 73a and 73b line by line, pixel by pixel, field by field, and screen by screen. It is written every time.

The subfield conversion table 710 is shown in FIGS. 3 (a) to 6 (a). As shown in these figures, the sub-field conversion table 710 shows that each video signal is time-sequentially arranged in the order of 1, 2, 4, 7, 13, 23, 33, 43, 5, 5.
This table shows the correspondence between the input signal for converting into the on / off information of the 10-bit subfields SF1 to SF10 having monotonically changing luminance weights of 5 and 74 and the combination of the converted subfields. The vertical column indicates the value of the input digital video signal (a ′), and the horizontal column indicates the 10-bit field information to which the input video signal should be converted. In these figures, the subfield marked "1" is "ON (lit)".
And other subfields mean that the field period is “off (non-lighting)” (hereinafter,
As well).

For example, in the subfield converter 71,
When a digital video signal having a value of 27 (* in the figure) is input, the video signal is converted into the subfield conversion table 7
Based on the table of 10, 1 called "0000011111"
Converts to 0-bit data and outputs. Note that the bit representation here is represented by associating the subfield number with the digit in the bit representation. By the way, if the converted 10-bit data is expressed by a decimal number, a value such as the value “31” described in the rightmost column in the figure is obtained.

Each of the frame memories 73a and 73b has an internal structure as shown in FIG. That is, the frame memory 73a stores the front half (1 to L (2
First field field information equivalent to 40 lines)
Memory area 73a1 and the front half of another screen (1
And a second memory area 73a2 for storing field information corresponding to L (240) lines. The frame memory 73b also includes the rear half of one screen (L + 1 to 2L (48
0) First field storing field information corresponding to line)
Memory area 73b1 and the rear half of another screen (L +
A second memory area 73b2 for storing field information corresponding to 1 to 2 L (480 lines).

Then, the first memory area 73a1 (first
Memory area 73b1) and the second memory area 73a2
The memory area of the (second memory area 73b2) includes 10 subfield memories SFM1 to SFM10.
Is equipped with. With this configuration, field information relating to a combination of 10-bit subfields corresponding to two screens divided into the first half and the second half for one screen is
It is written in the subfield memories SFM1 to SFM10 as information regarding lighting / non-lighting of each subfield. In the present embodiment, the subfield memory SFM1
~ SFM10 uses a semiconductor memory of 1-bit input and 1-bit output. In addition, this frame memory 73
a and 73b simultaneously write the field information and at the same time P
It is a 2-port frame memory that can read data to the DP 9 at the same time.

To write field information in the frame memories 73a and 73b, the first half of the field information for one screen is stored in the first memory 73a1 and the second half of the field information for one screen is stored in the first memory. 73b1 and the field information of the first half of the next one screen to the second memory area 73a2, and the field information of the second half of the other one screen to the second memory area 73b2. The four memory areas 73a1, 73b1, 73a2 or 73b2 of the frame memories 73a, 73b are alternately performed. Then, one memory area 73a1, 73b
The writing of the field information to 1, 73a2 and 73b2 is said to be performed by dividing the 10-bit data output from the subfield converter 71 in synchronization with the pixel clock into the 10 subfield memories SFM1 to 10 and writing the divided bit by bit. Executed in a way. Which bit of the 10-bit data is to be stored in which subfield memory SFM1-10 is predetermined.

Specifically, the subfield conversion table 7
The 10 subfield numbers 1 to 10 are logically associated with the subfield memories SFM1 to SFM having the same numbers as the subfield numbers, and the subfield numbers correspond to which subfield number the bit of the 10-bit data corresponds to. It is written in the subfield memories SFM1 to SFM10. 10-bit data subfield memory S
The write position to the FMs 1 to 10 is designated by the address designation signal from the write address controller 72. 10
Generally, the pixel signal before being converted into bit data is written at the same position as the position on the screen.

The display controller 8 includes a display line controller 80 and address drivers 81a and 81a as shown in FIG.
b and a line driver 82.

The display line control unit 80 has a memory area 7 to be read by the PDP 9 in the frame memories 73a and 73b.
3a1, 73b1, 73a2 or 73b2, a line and a subfield are designated, and an instruction as to which line of the PDP 9 is to be scanned is issued.

The operation of the display line control section 80 is synchronized with the writing operation to the frame memories 73a and 73b in the second encoding section 7 in the order of the screen unit. That is, the display line control unit 80 does not read from the memory areas 73a1 and 73b1 (73a2 and 73b2) in which 10-bit data is being written, but the memory area 7 that has already been written.
Reading is performed from 3a2, 73b2 (73a1, 73b1).

The address driver 81a outputs the subfield information corresponding to one line serially input bit by bit based on the memory area designation, the read line designation and the subfield designation of the display line control section 80. Bits (640 bits) corresponding to the number of pixels are converted into address pulses in parallel and output to the front half line of the screen. Similarly to the line driver 81a, the address driver 81b converts the subfield information into address pulses and outputs the address pulses to the lines corresponding to the second half of the screen.

The line driver 82 designates which line of the PDP 9 the subfield information is written to by a scanning pulse.

With the configuration of the display controller 8 as described above, field information is read from the frame memories 73a and 73b to the PDP 9 as follows. The reading of the field information for one screen that is divided and written in the frame memories 73a and 73b is performed by simultaneously reading the data corresponding to the first half and the second half. That is, the sub-field information SFM1 and SFM is obtained from the memory areas 73a1 and 73b1 at the same time for each pixel.
2, ..., SF10 are sequentially read out. More specifically, first, the memory area 73a
Subfield memories SFM1 to 1 of both 1 and 73b1
Subfield information corresponding to each pixel on the line is sequentially read bit by bit. And the line driver 8
After waiting for the line designation by 2, the latent image is formed (addressing) on the first line of each of the first and second half screens, and then the sub-field corresponding to each pixel of the second line of the first and second half screens from the same subfield memory SFM1. The field information is read and the address drivers 81a, 8a
Bits corresponding to the number of pixels in one line, 640 bits of subfield information in parallel, are output to PDP 9 in parallel for addressing. When such reading (writing) is completed up to the last line in each of the divided areas obtained by dividing the screen, the pixels are simultaneously emitted.

After the subfield information regarding lighting / non-lighting of the next subfield SF2 is read line by line in the same manner as described above and the addressing is performed, the operation is sequentially repeated until the subfield SF10.
The reading (writing) of the field information for the screen is completed.

FIG. 9 shows the operation method of such a PDP. In FIG. 9, the horizontal axis represents time, and the vertical axis represents the number of electrodes extending in the horizontal direction of the PDP, that is, the numbers of scan / discharge sustaining electrodes. The shaded areas indicate the addresses of the pixels to be illuminated and are shaded. The pixels are made to emit light in the part. In other words, addressing is performed by applying an address pulse to the address electrodes running in the vertical direction at the timing when the subfield SF1 starts for all the horizontal pixels on the scan / discharge sustaining electrodes of the first line of each divided screen. To do. After the addressing of the first line of the scan / discharge sustaining electrodes is completed, the same operation is repeated for the subsequent lines. When the last scan / discharge sustaining electrode addressing is completed in the split screen, the time t1 to t2 discharge sustaining period starts. During this period, the number of discharge sustaining pulses proportional to the weighting is applied to the discharge sustaining electrodes, but only the pixels that have been instructed to emit light by the above address designation emit light. Then, as will be repeatedly described, the gradation display for one TV field is completed by repeating the operations of addressing in the subfield and simultaneous lighting of all pixels as described above. Although not described here, the addressing is performed after an initialization period for erasing the wall charges of all pixels, and after writing information in advance to the pixels to be displayed in this way (addressing), The driving method of emitting light is called a "memory driving method".

Then, in parallel with the above-mentioned reading, the moving image is displayed by reading the field information corresponding to the first half and the second half of the next screen written in another memory area in the same manner as described above.

Next, the encoding characteristics of the second encoder 7 will be described.

In the subfield conversion table 710, the number of subfields is 10, which is shown in FIG.
~ As shown in Fig. 6 (a), 1, 2, 3, 4, 7, 13, 23, 33, 43, 55, which increase monotonically with time,
It was weighted with 74.

According to such weighting, the luminance of the subfield having a higher weight can be expressed by combining a plurality of subfields having a lower weight. As a result, there may be some combinations of subfields for performing the corresponding gradation display. For example, if the signal level is 127 (in the figure,
), Subfields SF10, SF8, SF4, SF
2, SF1 combination or subfield SF
There are combinations of 9, SF8, SF6, SF3, SF2 or subfields SF9, SF7, SF6, SF5, SF2, SF1.

Of these plural combinations, the subfield conversion table 710 is described by one of the combinations. That is, when the value of the digital video signal is 127, the subfields SF9, SF
7, the combination of SF6, SF5, SF2 and SF1 is described.

In this way, the subfield conversion table 7
In summary, the combination of subfields described in No. 10 is a combination in which use of subfields having higher luminance weights is suppressed as much as possible. Then, it is understood from the contents of the conversion table 710 that such a combination is adopted from medium brightness except low brightness (0 to 22) to high brightness (23 to 255).

Further, the signal level is 27 → 28, 50 → 5
As will be understood by referring to the combination of subfields at the positions changing from 1, 83 → 84, 126 → 127, 181 → 182, the following features are also provided. That is, at the level at which a subfield having a large luminance weight that was off up to a level lower by one gradation level (27, 50, 83, 126, 181) is turned on, the subfield is one stage lower than the luminance weight turned on. Subfields with low luminance weights are turned off.

By selecting the lighting pattern as described above, it is possible to prevent the lighting pattern from changing drastically with time when the signal level changes, which leads to elimination of the pseudo contour of the moving image.

Next, the characteristic contents of the invention will be described in detail.

First, as shown in FIG. 10, the filter unit 1 includes a two-dimensional high-pass filter 11, a two-dimensional low-pass filter 12, a temporal LPF 13 which is a time response low-pass filter, and an adding unit 14. It consists of and.

The two-dimensional high-pass filter 11 extracts only the fine picture component of the image. Of the extracted fine pattern components, the components that change drastically in the time direction are suppressed by the temporal LPF 13 and added by the addition unit 14.
Is output to.

The outputs of the temporal LPF 13 and the two-dimensional low-pass filter 12 are combined by the adder 14 and, finally, only the portion of the image components included in the input video signal where the fine image components drastically change in the time direction. It will be suppressed and displayed below. Therefore, the component in which the fine pattern changes in a short cycle is not displayed, and it is possible to prevent the noise component from being displayed. The high-definition image information of the still image portion, which is important for normal high-definition image display, is stored and displayed. Therefore, in a still image, fine image information is not impaired, and in moving image display, both a still image and a moving image are not impaired without impairing the response characteristics in a low spatial frequency portion such as a large area portion of the image. In this case, it is possible to display a good image with less noise.

The first encoder 6 removes the lower 4 bits of the 12-bit digital input video signal (signal (a '')) (signal (a ')) and converts it into an 8-bit signal (b). 3 (b) to FIG. 6 (refer to FIG. 3B to FIG. 6), in which the gradation display characteristics are sacrificed as the image motion amount increases and the pseudo contour elimination is prioritized. The input level (a ') is converted into a predetermined level (b) by referring to the conversion table 60 shown in b). It should be noted that 12
The gradation is displayed by using the upper 8 bits of the bits because the number of apparent gradations is increased.

This conversion table 60 converts the original level (a ') into one level (a') close to the original level (a ') based on the value of the output value m generated by the motion amount calculation unit, which will be described later, which indicates the amount of motion of the image. b) is a table showing correspondences for performing, and the column on the left end of the table shows the value of the input digital video signal (a ′) from which the lower 4 bits are removed. The column of (2) indicates a combination of signal values to be output with respect to the input signal at the corresponding motion amount m, that is, a mode of encoding the input signal. As will be described below, in the present image display device, it is necessary to delay the current input frame by one frame from the actual input video signal in order to calculate the motion amount prior to displaying the currently input frame on the PDP. Therefore, in the present embodiment, a frame memory (not shown) that can store at least two frames of video signals is provided, and the data is read from this frame memory and the first encoding unit 6 performs the encoding process.

The motion amount calculation unit 5 includes frame memories 51a and 51b for storing images for each one frame, a motion amount detection unit 52, an inclined portion detection unit 53, and a motion amount correction unit 5.
4 and.

The motion amount detecting section 52 includes a frame memory 51.
When a video signal is read from a and 51b, a frame to be displayed from immediately before, a previous frame, and a video signal for two frames are compared pixel by pixel, if the difference value exceeds a predetermined value, there is a movement (fluctuation). Here, the difference value is divided into 9 steps as an example, and the total is 10 including the detection that there is no movement.
A 4-bit value of "0000" to "1001" is output in stages. The larger the variation value from the previous frame, the more intense the movement (variation) of the pixel. The video data stored in the frame memory 51a (51b) is updated to the video data of the next one frame at any time after the display of the current one frame on the display is completed.

The inclined portion detecting portion 53 includes a frame memory 51.
The video signal is read from a (51b) and the edge portion in the same frame (pixel area where the signal level changes greatly)
Other than that, it is a circuit that detects an inclined portion that is an image area in which the level monotonously changes. Then, if an inclined portion is detected, "1" is added to the bits of the same value to make 4 bits,
If an edge part or a flat part with almost no level change is detected, "0" is added to the bit of the same value and output as 4 bits. In other words, if it is "1", then "1111"
If it is "0", "0000" is output. The slope detecting unit 53 uses a known edge detection filter to detect the inclination in the horizontal direction and the vertical direction, and if it is monotonically changing in any direction, the slope is detected.

The motion amount correction unit 54, the motion amount detection unit 5
The output from 2 and the output from the slope detection unit 52 are input, and finally, the amount of movement of the pixel from the previous frame is graded into 10 stages of "0" to "9", and the stage is adjusted according to the stage. The value m is output to the first encoding unit 6. Specifically, as shown in FIG. 12, the pixel does not move (value “000”).
In the case of “0”), the motion amount correction unit outputs “0000” regardless of the detection result of the inclined portion detection unit.
The pixel is in motion (values “0001” to “10
01 "), and if the signal level belongs to an area where the signal level monotonously changes between adjacent pixels, that is, a slope portion (when the value is" 1111 "), the signal is output in nine stages according to the amount of movement. (Values "0001" to "1001"). On the other hand, even if the output from the motion amount detecting unit is a value indicating that the motion is large, if it indicates that the output from the slope detecting unit 52 does not belong to a region that monotonically changes between adjacent pixels, (When the value is “0000”), the output value of the movement amount m is “00
00 ", that is, output as no motion. This does not cause a pseudo contour of a moving image depending on an image pattern such as an edge portion or a flat portion with almost no level change even if there is a motion. In this case, therefore, the first encoding unit 6 gives priority to the number of gradations. This is for encoding. Note that, in FIGS. 3B to 6B and the following description, the motion amount m is not a binary number, but is a decimal number for simplicity.

Returning to FIG. 11, the error diffusion unit 4 includes the addition unit 41.
The error calculation unit 42, the delay units 43a to 43d, and the coefficient units 44a to 44d.

The error calculator 42 is a circuit for calculating the difference (c) between the output level (b) as a result of the encoding in the first encoder 6 and the 12-bit input level (a ″). .

The delay unit 43a delays the difference value (c) by one pixel (1D) and outputs it, and the delay unit 43b calculates the difference value (1 line (1H) +1 pixel (1D)). Circuit for delaying and outputting a value, delay unit 4
3c is a circuit for delaying and outputting the difference value for one line (1H), and the delay unit 43d delays and outputs the difference value for (1 line (1H) -1 pixel (1D)). It is a circuit for.

The coefficient parts 44a to 44d are circuits for distributing the difference value (c) to a ratio corresponding to a predetermined coefficient, and the distributed value is finally displayed in the adding part 41 as the video signal of the pixel currently displayed. And output to the first encoding unit 6. Such processing is generally called an error diffusion method. Therefore, the input signal (a ″) input to the first encoding unit 6 is a signal obtained by adding the error of the neighboring pixel generated by the encoding process to the original signal level of the pixel. Is. The error diffusion process is performed regardless of whether it is a moving image or a still image. This is because the first encoding unit 6 cuts off the lower 4 bits of the 12-bit input video signal, so that an error of 4 bits always occurs even in a still image.

Next, the operations of the first encoding unit 6 and the error diffusion unit 4 will be specifically described.

First, FIGS. 3A to 6A and FIG.
As can be seen by comparing (b) to FIG. 6 (b),
The coding in the coding unit 6 has a correlation between the degree of change from non-lighting to lighting of a subfield and the amount of movement.

Specifically, in a combination of subfields, the subfields to be lighted have no change from non-lighting to lighting, are continuously lit from the beginning, and a combination of levels (“0”, which is most unlikely to cause pseudo contour).
"1", "3", "7", "14", "27", "5"
0 "," 83 "," 126 "," 181 ", and" 2 "
55 ") is used when the movement is the most intense (m =" 9 "). Then, in the display of one gradation lower, the non-illuminated subfield is proportional to the luminance weight of the subfield that changes to the illumination. The level is selected so that it causes a pseudo contour if the movement is intense, as occurs in. Another point of view is to select a signal level that has a long lighting time as the movement increases.

If the encoding is performed according to the degree of movement of such an image, the pseudo contour can be accurately eliminated when the movement is vigorous, while on the other hand, when the movement is small in which pseudo contour is unlikely to occur, the movement is small. , The number of gradations is increased and still images (m = “0”) can be expressed with the maximum number of gradations.

Specifically, in a portion where the input signal is almost a still image, the value of the motion amount (m) becomes "0", and the value shown in FIG.
The usable coded output (b) in (b) to (b) of FIG. 6 is 256 of the input signal (a ′) as shown by the ● mark in the figure.
Use signs of all kinds. That is, in the case of a still image, input (a ′) = encoded output (b), and image display is performed using 256 gradations.

On the other hand, as the motion of the image increases (the value of m increases), the set of usable coded outputs decreases. The maximum value of the motion amount (m) is "9", and when the maximum value is "9", the number of codes that can be used as the encoded output (b) is "0" or "0" as described above. 1 ",
"3", "7", "14", "27", "50", "8"
1 of 3 ”,“ 126 ”,“ 181 ”, and“ 255 ”
There is one type. This assigns weights W1 (1), W2
(2), W3 (4), W4 (7), W5 (13), W6
(23), W7 (33), W8 (43), W9 (5
5) and W10 (74), "0", "W1",
“W1 + W2”, “W1 + W2 + W3” ,. ． ． , "W
1 + W2 + W3 +. ． ． This means that the signal level is limited to (10 + 1) types of signal level of “+ W10”. Then, as the amount of movement decreases, W1, W2 ,. ． ． , WN is increased to increase the number of gray levels that can be expressed.

As the value to be limited for the input signal (a '), a level close to the (a') is selected. For example, in the case of “m = 9”, the signal level in the range of “1 to 2” is set to “1”, the signal level in the range of “3 to 6” is set to “3”, The signal level in the range of 7 to 13 "is" 7 ", and the signal level in the range of" 14 to 26 "is" 1 ".
4 ”, the signal level in the range of“ 27 to 49 ”is“ 27 ”
And the signal level in the range of "50-82" is "50",
The signal level in the range of "83-125" is "83",
The signal level in the range of “126 to 180” is “126”
In addition, the signal level in the range of "181 to 254" is "18
The signal level of "1" and "255" is limited to "255".

Therefore, for 11 kinds of light emission, for example, as the input level increases, the illuminated subfield pattern of the output level gradually extends (FIGS. 3B to 6B). ), The correlation between the magnitude of the input signal and the light emission pattern is guaranteed. In other words, since the subfield that was on at a low signal level does not disappear and remains on as it is, the distribution of the lighting pattern of the subfield can have a monotonic relationship that correlates with the signal level, and the signal level can be large. Then, the distribution of the light emission pulse spreads almost simply. Therefore, when an image is displayed using only such a limited light emission pattern, it is possible to eliminate a moving image pseudo contour peculiar to so-called moving image display.

The monotonic correlation between the distribution of the lighting patterns of the subfields and the signal level is a relationship that is substantially established even when the value of the motion amount m is "1" to "8". However, as the amount of movement decreases, the number of signal level selections increases, so things change slightly.
Lighting patterns at close levels will change slightly dramatically with time. When the amount of motion is small, the number of signal level selections is increased because the smaller the amount of motion, the greater the weighting of the luminance. This is because it does not contribute much to

By the way, in this case, if the movement is the most intense, an image is displayed with at most 11 gradations, and the number of gradations is obviously insufficient for displaying a natural image, and the pseudo contour is eliminated. Even if it is possible, the original image reproducibility is lacking. In order to eliminate this inconvenience, the error diffusion unit 4 having the above-mentioned configuration performs an error diffusion process. That is, the difference between the input level (a ″) and the input level that gives the limited encoded output (b) is fed back to the peripheral pixels as the error signal (c) to reduce the average error.
Specifically, as shown in FIG. 13, the pixel P currently being displayed is displayed.
, An error signal (c) is obtained, and this is distributed to four peripheral pixels, that is, A, B, C and D in FIG. The distribution coefficient is, for example, 7/16 of the error to the pixel A and 1/1 of the error.
Distribute 6 to pixel B, 5/16 of the error to pixel C, and 3/16 of the error to pixel D. The distributed error is encoded again by adding the error component to the original video signal. By repeating this, newly generated errors are distributed to the peripheral pixels one after another, and the average value of the display brightness almost matches the average brightness of the input, and the lack of gradation can be compensated.

It should be noted that the error diffused to the peripheral pixels is large in a portion having a lot of movement, and it is considered that the diffused error becomes conspicuous as noise. Is moved, so that so-called diffusion noise associated with such error diffusion processing can be displayed without much concern.

On the other hand, in the image portion determined to be a still image,
Since the encoded output (b) capable of displaying 256 gradations can be selected almost as described above, the diffusion noise is not observed. Further, the error diffusion process is 1
Since the calculation precision is 2 bits and this is also performed in the still image area, the effect that the apparent gradation in that area can be increased to 256 gradations or more can be expected.

[Second Embodiment] FIG. 14 is a block diagram of an image display device according to the present embodiment.
Differences from the image display device described in the section will be described. The image display device has a third encoding unit 101 in addition to the component filter unit 1, the γ inverse correction unit 2, the AD conversion unit 1, the display control unit 8, and the PDP 9 according to the first embodiment.
And a fourth encoding unit 102. The components having the same reference numerals as those of the image display device of the first embodiment shown in FIG. 1 have the same function.

The third encoding unit 101 is only for generating an 8-bit signal with the lower 4 bits of 12 bits removed, and here, the amount of motion as performed by the first encoding unit as described above is used. Is not encoded.

As shown in FIG. 15, the fourth encoding unit 102 sets the signal level represented by 8 bits obtained by removing the lower 4 bits of 12 bits in the third encoding unit 101 into 19 subfields (subfields). Field SF1 to subfield S
F19) is a circuit for converting into field information. The luminance weighting of the sub-fields here is 16, 16, 16, 1 in the order of time as can be seen from FIG.
6,16,16,16,16,16,16,16,1
No. 6,16,16,16,8,4,2,1 and the first subfield group consisting of 15 subfields with “luminance weight = 16” located at the beginning, and thereafter It can be classified into a second subfield group consisting of four subfields located. The sum of the brightness weights of the subfields of the second subfield group does not exceed the maximum brightness weight (16) of the subfields of the first subfield group (1).
5), a value that cannot be expressed by the brightness weights of the subfields belonging to the first subfield group (here, 1 to 1).
5) can be expressed by combining the second subfield group. This makes it possible to combine the light emission of the first sub-field group and the light emission of the second sub-field group to express the total luminance weight over the whole gradation without discontinuity with respect to the change in the value of the input signal. And

Each signal level is converted into field information consisting of a lighting pattern as indicated by "●". The 19-bit field information thus converted is displayed on the PDP 9 while being controlled by the display controller 8 as described above. In addition, in the level of "16-255",
Although SF16 to SF19 are combined to display the levels of 1 to 15, this lighting pattern is omitted for simplification.

FIG. 16 shows a PD in the image display device.
It is a figure explaining the light emission system of P9. In this case as well, as in the first embodiment, one screen is divided into two and driven by a method of simultaneously addressing vertically. As shown in FIG. 16, during one TV field, the initialization period R1
~ R5 are provided to initialize the charge state of the panel. This initialization corresponds to clearing all screens at once. After this,
By using the address period (denoted by symbol A), a voltage is selectively applied only to pixels to emit light to form so-called wall charges. The actual light emission is performed in the display periods D1 to D19. In addition, the numerical value in the parentheses added to the description of D1 to D19 in the drawing represents the brightness weight (hereinafter, the same).

As described above, in this case, the initialization corresponding to the erasing is performed only 5 times of R1 to R5. That is, the initialization period R1 is provided only before the address period of the subfield SF1 between the first subfield group of the subfields SF1 to SF15, and before the address period of the subfields SF2 to SF15. Has no initialization period. Therefore, with respect to the pixel whose light emission has started once, the wall charge remains held, and light is continuously emitted until immediately before the initialization period R2 after the end of the subfield SF15. On the other hand, for the second sub-field group consisting of four sub-fields having a small luminance weight on the rear side, the initialization periods R2 to R5 are provided prior to each address period, like the conventional driving.
The start and stop of light emission of each subfield are independently controlled.

With such an encoding and driving method, the larger the value of the input signal, the light emission starts at the head subfield position, and the subfield that emits light as the value of the input signal becomes forward. The encoding is such that the data is extended to (see arrow Y1 in FIG. 15). That is,
As described above, since the correlation between the magnitude of the input signal and the light emission pattern is guaranteed, it is possible to eliminate a moving image pseudo contour peculiar to so-called moving image display.

It should be noted that SF16 to SF1 having small luminance weights
In No. 9, the distribution of lighting / non-lighting changes irregularly to some extent, but since the luminance is small, the influence on the generation of the moving image pseudo contour is negligible.

Further, according to the present embodiment, although the total number of subfields is 19, the initialization period required for controlling the stop of light emission is only 5 times in one TV field period, The time required for initialization can be greatly reduced. Therefore, the total number of subfields can be increased in this way as compared with the conventional case.

Here, 2 PDPs with 480 lines are used.
Taking the case of line simultaneous driving as an example, one initialization period is 300 us, and an address period per line is 2 u.
If s, the display period for one TV field is (1
/ 60) × 1000000us- (300usx5 + 2
usx240x19) = 6000us. If the period of one light emission pulse in the display period is 5us, then 6000
It becomes us / 5us = 1200 times, and it becomes possible to emit light while ensuring sufficient brightness.

[Third Embodiment] The image display device capable of multi-gradation display according to the present embodiment has the same structure as the image display device according to the second embodiment except that the driving method is different. I will explain the points.

In the fourth encoding unit 102, the 8-bit signal is again used in 19 subfields (subfield SF
It is a circuit for converting into field information consisting of 1 to subfield SF19). The luminance weighting of the sub-fields here is, in order of time, as can be seen from FIG.
1,2,4,8,16,16,16,16,16,1
6,16,16,16,16,16,16,16,1
6 and 16, which are located at the beginning of "luminance weight =
It is classified into a second subfield group consisting of 4 subfields of 1, 2, 4, 8 "and a first subfield group consisting of 15 subfields of a luminance weight" 16 "located thereafter. be able to. That is, Example 2
The arrangement of the first subfield group and the second subfield group is interchanged with the case of. Then, each signal level is converted into field information composed of a lighting pattern as indicated by "●". The description of the lighting patterns of SF1 to SF4 is omitted for simplification.

FIG. 18 shows the PD in the image display device.
It is a figure explaining the light emission system of P9. As shown in FIG. 18, during one TV field, the initialization period R1 to
R5 is provided to initialize the charge state of the panel. In this initialization, the area simultaneous erasing is performed during the period indicated by R1 to R4, and the entire screen simultaneous writing is performed during the period indicated by R5. In the address period (denoted by A) in the second subfield group, a voltage is selectively applied to only the pixels to emit light to form so-called wall charges as in the conventional case. In the period (A), a voltage is selectively applied to the pixels to be extinguished, and the information for extinguishing the pixels is written to the pixel portions that do not need to emit light. Normally, an address pulse is applied to the pixels to be displayed to form charges, but the reverse is the case where all the pixels are displayed in advance by applying a pulse to all the pixels to form the charges, and the pixels not to be displayed This is a method of removing charges by selectively discharging. Incidentally, such an addressing method is described in detail in JP-A-6-186929.

By driving in this way, in the necessary initialization period, initialization for erasing the entire screen is performed four times (R1, R2, R3, R4) and initialization for simultaneous writing of the entire screen. Is only 5 times in total (R5), and by significantly shortening the time conventionally required for initialization, as the value of the input signal becomes large as described in the second embodiment. The encoding is such that the subfield that emits light extends backward in the time direction (arrow Y in FIG. 17).
2). That is, since the correlation between the magnitude of the input signal and the light emission pattern is guaranteed, it is possible to suppress the generation of the pseudo contour of the moving image.

[Fourth Embodiment] An image evaluation apparatus according to the present embodiment will be described in detail below. In the present embodiment, image evaluation is performed on the assumption of an image displayed by a subfield driving method used in a PDP or the like as an image display device with pulsed light emission. Although not shown in detail, the image evaluation apparatus according to the present embodiment is composed of a commercially available personal computer, and the hardware configuration of a general computer system, that is, a CPU,
It is composed of a memory (RAM, ROM), a keyboard, a hard disk, and a display monitor. It differs from a general computer system in that an image evaluation program unique to the present invention is stored in a hard disk device, and the CPU executes the program.

FIG. 19 shows the function of the image evaluation apparatus according to the present embodiment, which performs a simulation to evaluate the image quality when a moving image display of an image display apparatus with pulsed light emission such as a PDP is performed. It is a functional block diagram.

As shown in this figure, the image evaluation device is
A subfield information setting unit 2 for setting information for dividing the TV field into a plurality of subfields as described above.
01, a subfield coding unit 202 that converts a signal value of an input image into a subfield signal that is a 1-bit time series signal based on set subfield information, and virtually by the subfield coding unit 202. On the screen from the reference point setting unit 203 that sets one pixel as a reference point on the displayed virtual image, and the input motion vector (here, a vector indicating the amount and direction of movement of a specific image per unit time). The route calculation unit 204 that assumes the movement of the line of sight that follows the movement of the image within the predetermined period of time, and the time at which each light emission pulse is generated is calculated from the subfield order and the luminance weight set by the subfield information setting unit 201. The light emission pulse time calculation unit 205 that generates light and the line of sight when the light emission pulse is applied from the generation time of each light emission pulse and the calculated route. That position, that is line-of-sight position calculating unit 20 for calculating the position of the line of sight on the screen when there is a pulse emission
6 and pixels near the path through which the line of sight of the image signal converted into the subfield signal passes.
Based on the output of the neighboring pixel selecting unit 207, the neighboring pixel coefficient calculating unit 208 for calculating the calculation coefficient for the route neighboring pixel selected by the neighboring pixel selecting unit 207, the coefficient obtained by the neighboring pixel coefficient calculating unit 208 A coefficient multiplication unit 209 that performs a process of multiplying the light emission amount of the pixel selected by the neighboring pixel selection unit 207, and a light emission amount integration unit 2 that integrates the value obtained by the coefficient multiplication unit 209 over one TV field.
It consists of 10. The integrated value of the light emission amount within one TV field time obtained by the light emission amount integration unit 10 is output as an evaluation image.

The subfield information setting unit 201
In order to simplify the description, here, for example, once in the first subfield and twice in the second subfield according to the luminance weight of each subfield, as shown in FIG.
4 times in the 3rd subfield, 8 times in the 4th subfield, 16 times in the 5th subfield, 32 times in the 6th subfield, 64 times in the 7th subfield, 8th
In the subfield, subfield information of 128 times and a total of 255 times of pulsed light emission is set.

The subfield coding section 202 codes the input video signal based on the set subfield information. This encoding is performed by the table (corresponding to FIGS. 3A to 6A) showing the correspondence between the signal level of the input video signal and the combination of subfields.
According to.

FIG. 21 is a conceptual diagram visualizing the image evaluation method of the image evaluation apparatus of this embodiment, in which each quadrangle represents one pixel on the display screen.

In the route calculation unit 204, first, a pixel position (a pixel represented by a quadrangle P in the drawing) set by the reference point setting unit 203 is used as a base point, and a predetermined position on this pixel P (here , The upper left point P ′ of the pixel is the origin, and an XY coordinate system is formed. Then, from the motion vector (Vx, VY) represented by this XY coordinate system, a path K of the line of sight in one TV field is assumed. In FIG. 21, one TV field has four pixels on the right side and three pixels on the lower side (pixel P
From the pixel to the pixel Q). Note that, here, the line-of-sight path is calculated from the motion vector of the image on the assumption that there is a strong correlation between the motion of the image and the motion of the line of sight following the motion.

In the light emission pulse time calculating section 205, the time required for initialization, the time required for addressing, and
Since the time until the next pulse emission is known, the time at which each pulse emission is performed is calculated based on this.
The time when the initialization of the pixel P is started is set as the reference time and one pulse emission is performed, and the time is approximate to one point.

The line-of-sight position calculation unit 206 uses the light emission pulse time calculated by the light emission pulse time calculation unit 205 and the motion vector (Vx, VY) representing the motion of the image per unit time to generate the pulse K when there is pulse light emission. The upper line-of-sight position is calculated.

The neighboring pixel selection unit 207 assumes a predetermined region including the line-of-sight position calculated by the line-of-sight position calculation unit 206, and here, a region having the same area as one pixel of the image display device, for example, the display device. A quadrilateral region having the same shape as the display pixel is assumed, and pixels belonging to this region and performing pulsed light emission are selected as neighboring pixels at the screen position of the line of sight. For example, the pixels R1 to R4 included in a region corresponding to one pixel with the corner being the corner are selected as the neighboring pixels at the position represented by the coordinates Ki (x, y) in the figure.

The neighboring pixel coefficient calculation unit 208 calculates the area ratio of each pixel included in the quadrilateral region as a neighboring pixel coefficient.

In the coefficient multiplying unit 209, the coefficient calculated by the neighboring pixel coefficient calculating unit 208 is multiplied by the light emission amount of the pixel selected by the neighboring pixel selecting unit 207 as a weight, but it is obtained by one pulse light emission. The evaluation value at the line-of-sight position represented by the coordinates Ki (x, y) in the figure is obtained by adding the value obtained by multiplying the light emission amount by the coefficient for the neighboring pixels.

According to the evaluation in consideration of the light emission of the pixels in the vicinity of the line of sight, it is possible to obtain an evaluation image that is closer to the actual condition. That is, it is possible to reflect a phenomenon in which the visual acuity of a moving object is lower than the visual acuity of a still image, that is, a phenomenon of visual acuity of a moving image (so-called dynamic visual acuity).

The light emission amount integrating unit 210 obtains the light emission amount observed at the reference point P in one TV field by integrating the evaluation values thus obtained up to the position represented by the route end pixel Q '. The route end Q ′ is the position of the upper left corner of the pixel at the route end. Then, when the evaluation of this one pixel is completed, the reference point is set again and the same processing as described above is performed. By repeating this for the previous pixel, an evaluation image for one frame (1 TV field) is obtained.

Next, an example of the operation of the image evaluation apparatus having such a configuration will be described based on the flowcharts shown in FIGS.

First, waiting for the input of the image to be evaluated, and if the input is made (Yes in step S1)
s), lighting information of the subfield is created and stored in the hard disk (step S2). This subfield lighting information is associated with each pixel in the data structure shown in FIG. 25 (this FIG. 25 exemplifies the data structure). At the same time, the motion vector MV of each pixel is also stored in association with each other. In this table, the subscripts P (1,1) to P (n, m) represent the pixel positions of the evaluation image corresponding to the screen to be actually displayed in the horizontal and vertical positions. The source of the subfield lighting information is set in advance by the evaluator and written in the memory or the hard disk as a table corresponding to the above-described FIG. 3A to FIG. 6A.

Next, the pixel to be evaluated is set to the reference point (l
Is set as the number of reference points (the number of l = 1 to lmax) (step S3). This setting is performed based on the input from the evaluator's keyboard. Of course, all pixels may be set in advance as reference points.

Then, in step S4, l = “1” is set, and the following processing (steps S5 to S12) is executed for each reference point. As described above in step S5, X, which is the origin of the upper left corner of the reference point Pl (l = 1), is set.
-Y coordinate system is set (FIG. 21), and the position of each pixel is converted into this coordinate system.

Then, the motion vector MV of the pixel Pl (l = 1) is read out, and the line-of-sight path K between this motion vector MV and one TV field and this path end point Ql (l =
1, FIG. 21) is calculated (step S6).

At this time, each pulse emission time ti (i = 1, 1)
2 ..., 255) is calculated (step S7).

In step S8, i = “1” is set (step S8), and an evaluation area for one pixel centered on the line-of-sight position Ki (i = 1) is set (step S9).

FIG. 26 is a table showing the correspondence between the emission time ti and the subfield SF, which is stored in the hard disk.

Here, at time ti (i = 1), it is determined whether or not the pixels in this area emit light.
And are used to make a determination (step S10). In particular,
In FIG. 26, the subfield SF to which the time ti belongs is searched, and the searched subfield S in FIG. 25 is searched.
It is checked whether Fs is turned on by a pixel in the evaluation area. If it is turned on, it is turned on in FIG. 25 (denoted by ● in the figure)
The information is written. It should be noted that the coordinates translated from the pixel position represented by the XY coordinate system are the pixel positions in the original image. Further, the contents of FIG. 26 are updated whenever a source of new subfield lighting information indicating what luminance weight is used to divide one TV field is set, and the contents of FIG. 25 are set. Since it is generated based on the source of the field lighting information, this content is also changed when the source of the subfield lighting information is updated.

If light emission is to be performed (YES in step S10), the area ratio in the area of the pixel to emit light is calculated for each light emitting pixel with the total area area being 1 (step S11). Next, the light amount obtained by one pulse emission is multiplied by the area ratio, and the result is added to calculate the light amount Ai (i = 1) at the line-of-sight position Ki (i = 1) (step S12). Note that the line-of-sight position Ki (i = 1)
When the area ratio in the region is used as the calculation coefficient in the calculation of the light amount in, when the area of the region exceeds 1 pixel and even if all the pixels in the region actually exist, the total light amount of the 1 pixel Will not be considered. But,
As the evaluation area becomes larger, it is considered that the influence of the light emission of the peripheral pixels in a wider range is taken into consideration, and the evaluation accuracy decreases. Therefore, by setting the calculation coefficient of the neighboring pixels to be small in this way and processing to reduce the influence of light emission of the peripheral pixels, the evaluation accuracy can be maintained at the same level as when the evaluation area area is smaller. Can be expected.

Such processing is performed while incrementing the light emission time from the time t1 to the light emission time t2 to t255 (step S14) (i = imax (25 in step S13).
5) Determined by whether or not).

On the other hand, if NO in step S10, the pixels in the evaluation area do not emit light, so that the line-of-sight position Ki (i =
The light amount is not calculated in 1), but is incremented (step S14) and the same process as above is performed for the next light emission time t2.

The consideration amount Ai thus obtained is integrated to obtain the observed light amount at the reference point Pl (l = 1) (step S15). It is determined whether or not the observed light amount has been calculated for all the set reference points based on whether l = 1lmax (step S16), and if No in step S16, it is incremented (step S17) and returns to step S5 again and the same as above. Calculate the observed light intensity.

If all reference points have been calculated (YES in step S16), the integrated light amount of the corresponding pixel is replaced with the original signal level, and the synthesized image is displayed on the display of the computer system (step S18).
The evaluator observes the display result and judges the quality of the image.

In the above operation, the subfield lighting information is generated in advance for all the pixels in step S2, but this processing can also be performed when actually trying to integrate the light amount at each line-of-sight position. . That is, step S9
After setting the evaluation area with, the pixels that enter that area are found,
Pixels that contribute to the amount of light at the line-of-sight position are determined. At this stage, it is also possible to generate subfield lighting information of the pixel and check whether the corresponding subfield emits light.

As described above, according to the present embodiment, not only one pixel on the path through which the line of sight passes, but also a predetermined calculation is performed with respect to light emission from a plurality of pixels near the path through which the line of sight passes. Since inconsistencies such as subjecting the image to the evaluation are changed, the instability that the image of the evaluation result fluctuates significantly even if the image moves slightly is eliminated. Since it can be set arbitrarily, such as in the oblique direction, the image actually viewed by the observer can be reflected to enable stable image evaluation.

Further, when the magnitude of the motion vector is 0 (zero), it completely matches the original image, and it can be obtained that the image quality does not deteriorate in the still image. This matches the image quality when actually observing a still image.

Further, according to the image evaluation apparatus, an image equivalent to an image obtained by following the moving image on the screen with a camera having pixels such as a CCD camera as a device for observing the moving image. Can be calculated.
However, when trying to evaluate the image with a CCD camera, C
Since the CD camera must be repeatedly scanned at high speed in accordance with the movement of the image, it is actually difficult to perform reproducible evaluation. In that respect, according to the simulation of the image evaluation apparatus of the present embodiment, it is possible to perform evaluation with good reproducibility and high reliability.

[Other Matters] (1) In the first embodiment, the amount of motion is detected in 10 steps, but more simply, it detects only a still image or a moving image or only in a binary manner. The output can be limited to the signal level, and in the case of a still image, the input signal can be output as it is. It is also possible to detect the amount of motion in three stages such as vigorous / medium / absent, and devise the encoding as described above based on that.

Further, in the above 10 sub-fields, the respective luminance weights are 1, 2, 4, 7, 13, 23, 33,
Although 43, 55, and 74 are used, it goes without saying that the brightness weights are not limited to this, and the brightness weights are set to 1: 2: 4: 8: 16: 24: 32: 48: 5.
It may be 6:64.

Alternatively, with 12 subfields, 1: 2: 4: 8: 12: 16: 24: 28: 32:
The composition ratio of the brightness weights may be 36:44:48. In addition, the number of subfields is 11, and the luminance weights of the subfields are 1: 2: 4: 8: 16: 24: 32: 3.
It may be 6: 40: 44: 48.

Furthermore, the number of subfields is set to 9, and the luminance weights of the subfields are set to 1: 2: 4: 8: 16: 32: 4.
It may be 8:64:80.

Further, the respective brightness weights which have been generally used in the past and are apt to generate the pseudo contour pointed out in the conventional example are 1:
It may be eight subfields of 2: 4: 8: 16: 32: 64: 128. In this case, the number of signal levels limited to “present” and “absent” is changed. For example, when there is movement, the input signal level described in the leftmost column is changed to the rightmost column as shown in FIG. Generation of pseudo contours is suppressed by limiting to the signal level described, and when there is no movement, it is represented by the total number of gradations 0 to 255. Here, the degree of movement is divided into three stages of intense, medium, and none, and when the movement is intense, the signal level is limited as shown in FIG.
It is also possible to give priority to the number of gradations as the motion becomes smaller and to encode as in FIGS.

However, the larger the number of subfields, the smaller the change in the luminance weight, and the smaller the change in the distribution of lighting / non-lighting of the subfields. Therefore, the effect of suppressing the moving image pseudo contour is It seems to be more remarkable than when the number of subfields is small.

The order of forming these luminance weights may be in descending order. FIG. 3A in the case of this descending order
~ Figures corresponding to Fig. 6 (a) are shown in Figs. 30 to 33.

(2) The configuration of the filter unit 1 in the first embodiment is not limited to the above-mentioned configuration, and may be the configuration shown in FIG.

As shown in FIG. 34, the filter unit 1 includes a temporal HPF 301 which is a time response high pass filter,
Temporal LPF30 which is a time response low pass filter
2, a two-dimensional low-pass filter 303, and an adder 304
It can also consist of

By including the filter section having such a configuration, among the image components included in the input video signal, only the image component whose temporal change of the image is drastic is taken out by the temporal HPF 301. The two-dimensional low-pass filter 303 suppresses a part having a high spatial frequency component among the components of the part of the extracted image which has a large temporal change. The outputs of the two-dimensional low-pass filter 303 and the temporal LPF 302 are combined by the adder 304,
Eventually, among the image components included in the input image signal, the components having a spatially fine image component in a portion that changes drastically in the time direction are suppressed and displayed.

Therefore, similarly to the above, the component in which the fine pattern changes in a short cycle is not displayed, and the noise component can be prevented from being displayed. Further, since the high frequency component of the spatial frequency is maintained by performing the processing as described above, the response characteristic does not deteriorate when displaying a moving image, and the image quality such as the detailed portion of the image is not displayed. It does not cause a decline.

(3) In the first embodiment, the amount of motion is detected by taking the difference between frames for each pixel and detecting the variation value thereof. However, the present invention is not limited to this. By calculating the average variation value for each image block consisting of a set of multiple pixels,
A method by so-called pattern matching by matching with a template can be considered.

(4) The luminance weights of the sub-fields in the second and third embodiments are not limited to the above-mentioned configuration, but "23, 22, 21, 20, 19, 19, 18, 17, 1".
6,15,14,13,12,11,11,10,6
The luminance weights of the first 14 subfields of "4, 2, 1" are set to relatively large values (23, 2
2,21,20,19,18,17,16,15,1
A total of 19 subfields (4, 13, 12, 11, 11) (the first 14 subfields are set as the first subfield group, and the other subfields are set as the second subfield group). .) Or "24,
24, 24, 24, 24, 24, 24, 24, 24, 1
A total of 15 subfields in which the luminance weights of the top 10 subfields of "6,10,6,4,2,1" are "24" and "16" (the set of the top 10 subfields is the first A subfield group and a set of other subfields may be a second subfield group.). Although not described in detail in these cases, the sum of the brightness weights of the subfields of the second subfield group does not exceed the maximum brightness weight of the subfields belonging to the first subfield group. Values that cannot be expressed by the brightness weights of the subfields belonging to the first subfield group can be expressed by combining the subfields of the second subfield group.

Further, if the initialization is performed a smaller number of times than the number of subfields forming the first subfield group, the effect of increasing the display period and reducing the occurrence of pseudo contours can be obtained. it can.

Furthermore, if the weights of the subfields of the second subfield group are finely divided and the number of initializations is reduced,
It is possible to improve the image quality at low brightness.

(5) In the first to third embodiments, the AD conversion is performed after the γ inverse correction of the input analog video signal is performed.
The present invention is not limited to this, and it is also possible to perform γ inverse correction after performing AD conversion.

(6) By using the evaluation device of the fourth embodiment, a very good guideline is provided for the design of the image display device represented by the PDP, and the number of subfields, the respective brightness weights, etc. are determined. Easier and also
The resulting image display device also has fewer pseudo contours than the conventional one.

Further, the image evaluation function of the image evaluation apparatus is realized by a program for executing each of the above-mentioned functions, which is implemented by a floppy disk, an IC card, a ROM.
It can be easily implemented by an independent computer system by recording it on a recording medium such as a cassette and transferring or transferring it.

(7) Finally, the techniques of the above-described first to fourth embodiments can be similarly applied to a DMD (digital micromirror device).

[0147]

As described above, the brightness weights of the sub-fields of the present invention are W1, W2 ,. ． ． , WN, 0, W1, W2 ,. ． ． , WN that can be expressed in any combination, and a selection unit that selects one gradation value according to the amount of movement of the input video signal and a sub that expresses the selected one gradation value. According to the image display device provided with the subfield lighting means for lighting the field, the above-mentioned first object is achieved.

Further, the same purpose is to convert the input video signal into ON / OFF information of a plurality of subfields in pixel units, a display in which each pixel on the display screen is composed of a light emitting cell, and a conversion unit. One TV field of ON / OFF information converted by the means is distributed for each subfield, and the subfields are sequentially switched to turn on / off each light emitting cell of the display. Before lighting the subfield, And a display control unit for performing the initialization for a number of times equal to or less than (the number of subfields-1), wherein the conversion unit stores ON / OFF information of the plurality of subfields corresponding to each level of the input video signal. In the range of a predetermined input video signal, the on / off information indicates that the subfield that emits light in proportion to the gradation value of the input video signal is in the time direction. Square or ON as will extend rearwardly, also achieved by the image display apparatus is off information.

A second object is to hold information on subfields constructed in the device to be evaluated, convert an input video signal by the subfield information, and turn on which subfield for each pixel. Related to the video signal, a sub-field lighting information creating means for creating lighting information, a reference point setting means for setting one pixel as a reference point on a virtual image virtually displayed by the sub-field lighting information. According to the input motion vector, the route estimation means for assuming a route moving from the reference point in a unit time, and the light emission amount of the pixels existing around the movement position at each moment in the unit time as the subfield lighting information. The light emission amount calculating means obtained from the above, the light emitting amount at each moving position, the integrating means for integrating the light emitting amount over the moving route of the unit time, and the product It is achieved by the image evaluation apparatus including a evaluation means for obtaining evaluation information of the image display state of the target device from the value.

The third object is achieved by an image display device having a display and a filter means for suppressing the time response of the high frequency component of the spatial frequency component of the input video signal.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of an image display device according to an embodiment.

FIG. 2 is a block diagram showing a configuration of a second encoding unit 7.

FIG. 3 is a chart showing an aspect of encoding in the image display device. (A) is a chart which shows the aspect of a 2nd encoding part, (b) is a chart which shows the aspect of a 1st encoding part.

FIG. 4 is a chart showing an aspect of encoding in the image display device. (A) is a chart which shows the aspect of a 2nd encoding part, (b) is a chart which shows the aspect of a 1st encoding part.

FIG. 5 is a chart showing an encoding mode in the image display device. (A) is a chart which shows the aspect of a 2nd encoding part, (b) is a chart which shows the aspect of a 1st encoding part.

FIG. 6 is a chart showing an aspect of encoding in the image display device. (A) is a chart which shows the aspect of a 2nd encoding part, (b) is a chart which shows the aspect of a 1st encoding part.

FIG. 7 is a diagram showing a configuration of a frame memory of the image display device.

FIG. 8 is a block diagram showing a configuration of a display control unit of the image display device.

FIG. 9 is a diagram illustrating a light emitting method of a PDP in the image display device.

FIG. 10 is a block diagram showing a configuration of a filter unit of the image display device.

FIG. 11 is a block diagram showing a configuration of an error diffusion unit and a motion amount calculation unit of the image display device.

FIG. 12 is a chart for explaining generation of an output signal of a motion amount calculation unit of the image display device.

FIG. 13 is a schematic diagram for explaining a method of error diffusion of the image display device.

FIG. 14 is a block diagram showing a configuration of an image display device according to another embodiment.

FIG. 15 is a chart showing an encoding mode of a fourth encoding unit of the image display device.

FIG. 16 is a diagram illustrating a light emitting method of a PDP in the image display device.

FIG. 17 is a table showing an encoding mode of the fourth encoding unit of the image display device of yet another embodiment.

FIG. 18 is a diagram illustrating a light emitting method of a PDP in the image display device.

FIG. 19 is a functional block diagram for explaining the functions of the image evaluation apparatus of another embodiment.

FIG. 20 is a diagram illustrating a light emission pattern of an image used for simulation in the image evaluation apparatus.

FIG. 21 is a schematic diagram for explaining an image evaluation method in the image evaluation apparatus.

FIG. 22 is a flowchart showing an example of the operation of the image evaluation apparatus.

FIG. 23 is a flowchart showing an example of the operation of the image evaluation apparatus.

FIG. 24 is a flowchart showing an example of the operation of the image evaluation apparatus.

FIG. 25 is a chart showing a data structure when a pixel and subfield information of the pixel are stored.

FIG. 26 is a chart showing correspondences between light emission times and subfields.

FIG. 27 is a chart showing another aspect of the first encoding unit.

FIG. 28 is a chart showing another aspect of the first encoding unit.

FIG. 29 is a chart showing another aspect of the first encoding unit.

[Fig. 30] Fig. 30 is a table showing the encoding mode of the second encoding unit corresponding to Figs. 3 (a) to 6 (a).

FIG. 31 is a chart showing an encoding mode of a second encoding unit corresponding to FIGS. 3 (a) to 6 (a).

FIG. 32 is a chart showing an encoding mode of a second encoding unit corresponding to FIGS. 3 (a) to 6 (a).

FIG. 33 is a chart showing an encoding mode of a second encoding unit corresponding to FIGS. 3A to 6A.

FIG. 34 is a block diagram showing another configuration of the filter unit in the modification of the first embodiment.

FIG. 35 is a diagram for explaining the conventional image display device, and is a diagram showing a state in which a predetermined image pattern moves in parallel by two pixels.

FIG. 36 shows a state where the observer follows the state where the image pattern moves in parallel, and is observed.

FIG. 37 is a diagram for explaining still another conventional image display device, and a diagram corresponding to FIG. 36.

[Explanation of symbols]

1 Filter section 2 γ inverse correction unit 3 AD converter 4 Error diffusion section 5 Motion amount calculator 6 First Encoding Unit 7 Second encoding unit 8 Display control unit 9 PDP 11 Two-dimensional high-pass filter 12 Two-dimensional low-pass filter 13-hour response low-pass filter 13 14 Adder 41 Adder 42 Error calculator 43a to 43d delay unit 44a to 44d coefficient part 51a, 51b Frame memory 52 Motion Detection Unit 53 Slope detection unit 54 Motion amount correction unit 60 conversion table 71 Subfield converter 72 write address controller 73a, 73b frame memory 80 Display line controller 81a, 81b address driver 82 line driver 101 Third Encoding Unit 102 fourth encoding unit 201 Subfield information setting section 202 subfield coding unit 203 Reference point setting section 204 Route calculator 205 Emission pulse time calculation unit 206 Line-of-sight position calculation unit 207 Neighborhood pixel selection unit 208 neighborhood pixel coefficient calculation unit 209 coefficient multiplication unit 210 Light emission amount integration unit 710 Subfield conversion table

─────────────────────────────────────────────────── ─── Continued Front Page (51) Int.Cl. ⁷ Identification Code FI G09G 3/28 G09G 5/00 510S 3/36 5/36 510M 5/00 H04N 5/205 510 5/66 A 5/36 510 101B H04N 5/205 17/04 Z 5/66 G06F 15/62 340Z 101 G09G 3/28 K 17/04 5/00 520J // G06T 7/20 G06F 15/70 410 (31) Priority claim number Hei 9-333863 (32) Priority date December 4, 1997 (Dec. 4, 1997) (33) Priority claiming country Japan (JP) (31) Priority claim number Japanese Patent Application No. 9-341116 (32) Priority date December 11, 1997 (December 11, 1997) (33) Priority claiming country Japan (JP) (56) References JP-A-10-31455 (JP, A) JP-A-10-39830 ( JP, A) JP 5-127612 (JP, A) JP 7-302061 (JP, A) JP 11-231831 (JP, A) JP-A-11-296132 (JP, A) (58) Fields investigated (Int.Cl. ⁷ , DB name) G09G 3/20 660 G09G 3/20 641 G06T 13/00 G09G 3/28 G09G 3/36 G09G 5/36 510 H04N 5/205 H04N 5/66 101 H04N 17/04 G06T 7/20

Claims (6)

(57) [Claims] 1. An image display in which one TV field is configured by arranging N sub-fields each having a luminance weight in time order, and a desired sub-field is turned on to display an image of one TV field in multiple gradations. A device , comprising a motion detecting means for detecting a motion amount of an input video signal, and a motion amount of the input video signal detected by the motion detecting means,
Supports two or more stages of no motion and motion
Select a coded output set according to the amount of motion of the input video signal from multiple coded output sets
Then, the gradation of the input video signal is selected from the selected set of encoded outputs.
A display signal coding means for selecting a coded output close to the value and a subfield of the subfield corresponding to the selected one coded output.
And a subfield driving means for controlling lighting, wherein the display signal coding means has a large motion amount of the input video signal.
As the brightness increases,
Coded output that suppresses non-lighting / lighting changes in
Image display device characterized by selecting a combination of . 2. The display signal encoding means increases the number of codes as the amount of motion of an input video signal decreases.
The image display apparatus according to claim 1, characterized in that the selection of the set of encoded output consisting of the output. 3. The display signal encoding means is W1 ≦ W2 ≦ when the input video signal has the largest motion amount.
... ≤ WN, "0", "W1", "W1"
+ W2 "," W1 + W2 + W3 ",. ． ． , "W1 + W
2 + W3 +. ． ． + WN " encoding of N + 1 types
The image display apparatus according to claim 1 or 2, characterized in that selecting a set of coding output and an output. 4. The one having the largest corresponding motion amount among the sets of coded outputs , when the gradation value of the display signal becomes large.
The image display apparatus according to claim 1 , wherein the image display apparatus is a combination of coded outputs that satisfies a relationship in which the distribution of illuminated subfields simply spreads . 5. The image display device according to claim 1, wherein a gradation value of an input video signal is used for display for one pixel .
When the gradation value is different, according to claim 1-4, characterized in that it comprises an error diffusion means for distributing the difference to neighboring pixels
The image display device according to any one of 1. Wherein said error diffusion means, and around the error calculating unit for calculating an error tone value of the tone values used in the display and the gradation value of the input video signal, an error tone value signal the calculated A delay unit that delays to disperse to a predetermined pixel, a coefficient unit that determines a gradation value to be distributed to the pixels to be dispersed, and a gradation value that is distributed to each pixel obtained by the coefficient unit in the input video signal. The image display device according to claim 5 , further comprising: an addition unit that adds