JPS63171086A - Television signal processing circuit - Google Patents

Abstract

PURPOSE:To obtain a reproduced picture with high precision, which has small quantity of the deterioration of the reproduced picture even with the movement of the picture, by detecting the movement information of the picture of a TV signal and conducting an optimum interpolation signal with the aid of the detected signal. CONSTITUTION:An input TV signal is delayed in field memories 14 and 15 by a specified H respectively and the delay signal X and the input signal Z are subtracted 19, then the difference signal of (Z-X) is converted into a signal expressing the movement of a picture element Y in a coefficient circuit 20. The output from the memory 14 is the signal of a scanning line (n), which is delayed only by 1H in a line memory 16 so as to obtain the signal of the scanning line l. The mean value between the scanning line (n) and l is obtained in an addition circuit 17 and a coefficient circuit 18 and sets it as an interpolation value (a) from the signal in a same field. And the value Y of the picture element which should be interpolated is obtained by using the value (a), the interpolation value (b) from the former field and the movement coefficient (k) obtained in the circuit 20. The signal of the scanning line (n) and the signal of a new scanning line (m) consisting of the interpolation signal Y are inputted in a time base conversion circuits 24 and 25 and the output is switched in a 1/2 H period so as to double desirable scanning line.

Description

[Detailed Description of the Invention] 1. The present inventor's television signal processing circuit, more specifically, receives a currently broadcast television signal and generates a television image with twice the number of scanning lines by interpolation. It relates to a signal processing circuit to be obtained.

Currently, television receivers with larger screen sizes are being developed. When the screen size is enlarged, the scanning line spacing of the displayed television image naturally decreases, and the resolution of the image becomes insufficient, making it impossible to obtain high quality images. In addition, there is an increasing demand for obtaining high-quality images that are even finer than those of conventional television receivers.

In response to these demands, high-definition televisions with double the number of scanning lines are being developed. In this case, a high-resolution, high-quality television image can be obtained, but changing the number of scanning lines from imaging requires changing the existing transmitter and receiver, so it is difficult to put it into practical use. There are many technical and economic issues to be solved, and it will take a long time.

On the other hand, a high-resolution television receiver is being considered that receives signals from currently being broadcast television and interpolates between the scanning lines, thereby essentially doubling the number of scanning lines.

That is, in order to double the number of transmitted scanning lines, the received signal is subjected to time compression processing, and the scanning lines spanning two fields are made into a single frame signal. The NTSC television signal currently broadcast in Japan is sent by interlace scanning a field signal with 262.5 scanning lines every 1/60 seconds, and a frame signal with 525 scanning lines in 2 fields. configured. Therefore, when image signals of two fields that differ in time by 1/60 seconds are combined to form a single frame, if the image is stationary or has little movement, the number of scanning lines will be doubled on the high-definition side. However, if the image changes over time, 1/60
There is a problem in that the image is a composite of two images that moved for a second, which deteriorates the quality of one image. To solve this problem,
A field delay output, a line average output, and a switching circuit are prepared, and the presence or absence of image movement is determined and the switching circuit is controlled.Still parts are interpolated from the field delay output, and video parts are interpolated from the line average output of the same field. A method of switching to and interpolating is known from Japanese Patent Laid-Open No. 138325/1983. However, switching based on the presence or absence of movement causes a new problem in that unnatural image quality deterioration occurs at the switched boundary due to the difference in image quality between interpolation modes.

Therefore, one object of the present invention is to provide a television receiver that receives an interlace-scanned television signal and pseudo doubles the scanning lines inside the receiver, so that there is little deterioration of the reproduced image even when the image moves. The object of the present invention is to realize a television signal processing circuit that obtains high-definition reproduced images.

In order to achieve the above object, the present invention detects motion information of an image of the television signal when creating an interpolated scanning signal from a received television signal, and uses the detected signal to select an optimal interpolated signal from a plurality of types of interpolation means. It is characterized by being structured to obtain the following.

That is, a detection circuit that obtains image motion information from an interframe difference signal of an interlaced television signal, a first interpolation circuit that uses scanning lines within a field, and a detection circuit that obtains image motion information from an interframe difference signal of an interlaced television signal, a first interpolation circuit that uses scanning lines within a field, and a detection circuit that uses at least a previous field (or a previous and subsequent field). a second interpolation circuit that uses scanning lines, and a mixing ratio in which the mixing ratio of the output of the first interpolation circuit and the output of the second interpolation circuit is controlled by the output of a detection circuit that obtains motion information of the image. and a time-base conversion circuit for compressing the time-base of the mixed output to 1/2, thereby obtaining a television signal with twice the number of scanning lines of the input television signal. It is a processing circuit.

Hereinafter, the present invention will be explained in detail with reference to the drawings.

FIG. 1 shows a general configuration of a television receiver in which the scanning lines of an input television signal are artificially doubled by interpolation.

Regarding the current NTSC television broadcast signal as a signal source, a television signal with 525 scanning lines, a horizontal frequency of 15.75 kHz, a vertical frequency of 60 Hz, and a 2:1 interlace is a normal NTSC television broadcast signal. A demodulating device 1 similar to a television receiver obtains a Y signal, a ■ signal, a Q signal, a horizontal synchronizing signal H1, and a vertical synchronizing signal V.

Y.I.

Each signal of Q is input to the time axis conversion circuit 2 as explained in FIG.
The signal is converted into I' and Q' signals and applied to the video amplification circuit 4. Also, the H and V signals are converted to H' by the synchronous signal converter 3.

V' and applied to the display device 5 as a drive signal of the television signal with the scanning line doubled.

FIG. 2(A) shows the conventionally known time axis conversion circuit 2.
FIGS. 2A and 2B are diagrams illustrating the states of scanning lines of an input television signal and an interpolated television signal, respectively. The time axis conversion circuit 2 shown in FIG. 1 includes circuits as shown in FIG. 2(A) for each of the Y, I, and Q signals.

The Y signal will be explained below to simplify the explanation.

A Y signal corresponding to the current television signal 1 interlaced every 1/60 seconds is applied to the input terminal 6 as shown in FIG. A portion of the signal is delayed by one field period (1/60 second) by the field memory 7, and is input to the time axis conversion circuit 8.9 together with the input signal. The time axis conversion circuits 8 and 9 convert the signal into a signal whose time axis is compressed to 1/2. These output signals are converted to 1/2H by the switching circuit 10.
By switching each time, a time-base converted signal with a horizontal scanning period of 1/2H is obtained at the output terminal 11. Therefore, in the reproduced image, the horizontal scanning lines are doubled (
525) television images are obtained every 1/6o seconds. Further, if horizontal and vertical drive signals are generated so that the second field and the third field are interlaced, an image with 1050 horizontal scanning lines in one frame (two fields) is obtained.

However, among the 525 scanning lines of each field in the figure (C), half of the 262t5 scanning lines are from the previous field, that is, 1
/Since it is a pseudo signal created by interpolating the image signal from 60 seconds ago, if the image (subject) is moving or has many temporal changes, it is a composite of the previous field image and the current field image. Therefore, when there is a change in the image (image movement, brightness, color change, etc.) for 1/6o seconds, the image quality deteriorates.

In order to explain the principle of the television signal processing circuit according to the present invention, FIG. 3 is a perspective view showing the states of three consecutive fields of scanning lines of a current television signal. In the figure, the broken line in the first field, the dotted line in the second field, and the solid line in the third field are actually transmitted signals. From these signals, for example, a pseudo scanning line signal m is generated by interpolation between the scanning lines Q and n of the second field.
When creating an image, depending on the motion information of the image, the second field's scanning lines α and n may be used, or the first scanning line at the same position as m may be used.
Alternatively, the signal of the scanning line of the third field may be used. More specifically, when creating a pixel Y on a scanning line m created by interpolation, when the subject is moving, the upper and lower scanning lines Q of the same field are most closely approximated in terms of time.

Interpolate from n. In addition, if the subject is close to a still image with little movement, the closest one in the spatial domain, i.e., pixels X and Z at the same position as the above pixels in the first and third fields.
Use.

Note that if the subject is moving, the resolution will in principle deteriorate because pixels from different areas are used, but human vision has a characteristic that the resolution decreases for moving objects. Therefore, the reduction in resolution due to the interpolation described above does not substantially pose a problem.

FIG. 4 is a block diagram showing the main part of an embodiment of an apparatus implementing the television signal processing method according to the present invention, that is, the structure of a time axis conversion circuit. The television signal (for example, luminance signal Y) inputted from the input terminal 12 is stored in the field memories 14 and 15, respectively.
It is delayed by 21H and 263H (H is the horizontal scanning period). If the input signal is the pixel Z (of the third field in FIG.
signal level is represented by Z), field memory 15
The output is X (the signal level is represented by X) of the first field exactly one frame before. The signals X and 2 are applied to the subtraction circuit 19 and converted by the coefficient circuit 20 into a (Z-X) difference signal, that is, a signal representing the movement of the pixel Y in the second field.

On the other hand, the output of the field memory 14 is the scan in signal of the second field, which is sent to the line memory 16 by IH.
, and obtain the signal of scanning line 0. The average value of the scanning line n and Ω is obtained from the adding circuit 17 and the coefficient circuit 18.

Let this be the interpolated value a from the signal within the same field.

This interpolated value a.

Using the interpolation value b from the previous field and the motion coefficient k (0≦≦1) obtained by the motion coefficient circuit 20, the value Y of the pixel to be interpolated is set to Y=−b+(17k)・a・...Find as (1). The multiplication circuit 21 and the addition circuit 23 are circuits for calculating the above equation.

The signal of the scanning line n of the second field and the signal of the new scanning line m consisting of the interpolated signal Y obtained above are input to the time axis conversion circuit 24.25, and the output thereof is converted to 1 by the switching circuit 26.
By switching at the /2H period, a high-definition television signal with twice the desired number of scanning lines can be obtained at the output terminal 13.

In the above embodiment, the still image interpolation value is the signal of the previous field, but it can also be the average value of X and 2 using the addition circuit and the 1/2 coefficient circuit. In this way, minute noise components cancel each other out, making it possible to obtain a higher quality image.

FIG. 5 is a diagram showing the configuration of another embodiment of the television signal processing circuit according to the present invention, particularly the interpolation scanning signal generating section which constitutes the main part thereof.

In the embodiment shown in FIG. 4, the scanning method after doubling the scanning lines is such that one frame has 525 scanning lines, a frame signal is generated every 1/60 second, and no interlacing is performed. The embodiment shown in FIG. 5 further performs interlacing. Components in FIG. 4 are designated by substantially the same numbers as those in FIG. 4, and detailed explanation thereof will be omitted.

In this embodiment, the output signal of the field memory 14 (signal to be interpolated) and the signals one field before and after this signal are obtained by the adder circuit 17 and the coefficient circuit 18, and the average value signal is obtained by using the former directly or by IH. The latter is applied to the coefficient circuit group 30 either directly or via line memories 26, 27, and 28 that obtain delays of LH and 2H. The coefficient circuit group multiplies the value in the O mark for reasons described later. The output of the coefficient circuit is outputted directly or via the adder circuit group 31 and further through the changeover switches 32, 33, 34, 35 to the adder circuits aj, 3.
7 or added to subtraction circuits 38 and 39. The output of the subtraction circuit passes through the multiplication circuits 40 and 41 and is then sent to the addition circuit 36.
Added to 37.

The way the adder circuits 36 and 37 are output is the time axis conversion circuit 2.
4.26, the time axis is compressed to 1/2.

The changeover switch 26 passes through a coefficient circuit 42 for adjusting the coefficient multiplied by the coefficient circuit group 30 described in the time axis conversion circuit 24, and then becomes an interlaced signal with the number of scanning lines doubled. In addition.

When the time axis conversion section is constructed from a digital circuit, a D/A conversion circuit is added to the output side of the coefficient circuit 42. The operating principle of the above embodiment will be explained with reference to FIGS. 6, 7, and 8. FIG. 6 shows, on the same drawing, the positional relationship of the scanning lines of a current television signal and a television signal whose number of scanning lines has been converted and interlaced according to the embodiment shown in FIG. Solid lines, broken lines, and dotted lines indicate the positions of the scanning lines. Current television signals are composed of a first field (solid line ○ mark) and a second field (dashed line Δ mark), and the scanning line signal in the above embodiment is composed of the first field (solid line ○ mark) and the second field (dashed line Δ mark). The fields shown above are the solid lines and dashed lines (which are the scanning lines of the current television signal).
The second field scans the position of the dotted line (indicated by the square mark) that is interlaced between the scanning lines of the first field.

As can be seen from FIG. 5, in this embodiment, the interpolation operation changes depending on the motion information (circuits 19, 20, 41, 40, etc.), and the coefficient circuit group 3o, adder circuit group 31, etc. Efforts are being made to improve the frequency characteristics of the signal.

In the vertical direction of the screen, one image can be considered to be sampled by scanning lines. If the number of scanning lines is taken as a sampling frequency unit in the vertical direction, the current television signal has a sampling frequency of 525 lines, of which the sampling frequency is 262.5 lines for field images. That is, from the discussion of the previous embodiment, it can be considered that the sampling frequency of a still image of the current television signal is 525 lines, and that of a moving image is 262.5 lines. Furthermore, the sampling frequency after scanning line doubling (with interlacing) is 1050 lines. The scanning line doubling process is equivalent to the problem of designing a processing circuit that interpolates between sample points in the vertical direction. Figure 7 shows the sampling frequency f in the vertical direction of the image.
V is expressed on the horizontal axis using the number of scanning lines as a unit, and is 26
2.5 and 525 are field images (moving images), respectively.
and the sampling frequency of the frame image (still image), 10
50 indicates the sampling frequency fs after conversion. therefore,
The interpolation circuit constitutes a kind of interpolation filter, and out of the band 0 to 525, the pass band is (sampling frequency before conversion)/2, and the signal is generated around the a main frequency and frequencies that are integral multiples thereof. It can be designed as a low-pass filter (LPF) that suppresses harmonic components. In other words, in the case of moving images, the sampling frequency is the scanning line frequency within the field (
262,5), it becomes an LPF that passes vertical frequency components of 172 or less, and similarly for a still image, it becomes an LPF that passes 172 or less of the scanning line frequency (525 lines) in the frame. Curve Hm(t”) in FIG.
Hs(f) indicates the frequency characteristic of the interpolation filter,
Hm (f) is the LPF for moving images.

Hs(f) is the LPF for a still image. Its transfer characteristic is expressed by the following equation.

......(2) 2 π The impulse response of a filter with the frequency characteristic expressed by equation (2) is expressed as (1, 2, 3, 4, 3,2,1)716. This is an interpolation filter that linearly interpolates sample values. In this video processing mode, only signals from the same field can be used, so as shown in the video mode in Figure 8, there is a sample value of the input signal every four sampling periods of the impulse response, so in the first field, the solid line Interpolates the signal from the scanning line indicated by (O mark) to the scanning line marked by .
In the field, the scanning line indicated by the dashed line (Δ) is interpolated from the scanning line indicated by the mu mark using the coefficient (6, 2) shown in the figure.

In still image mode, all scan lines of the input signal are available for interpolation. The impulse response of equation (3) is (-1, -1,
5, 10, 2, -1, -1)/16, so in the first field, from the scanning line shown by the solid line (○ mark) and by the scanning line shown by the broken line (Δ mark), the mark is drawn from the scanning line, and in the second field, the mark is By interpolating the signal marked with mu from the input scanning line (marked 0, Δ) with the coefficients shown, a signal having the frequency characteristics shown in FIG. 7 is obtained. That is, in FIG. 5, the vertical position in FIG. 8 is determined by field memories 14 and 15 and line memories 27, 28° 29,
33, 34, and 35 alternately (every 1/60 seconds) switch the first field and the second field in the moving image mode and still image mode shown in FIG. , coefficient control circuit 20
The coefficients of the coefficient circuit 40.41 are given by and (1-k)
This is done by changing the image element by picture element.

In this still image mode scanning line interpolation circuit, the gain of the vertical high frequency component is larger than that of direct current, and an image with higher definition can be displayed.

In the above description, a television system with 525 scanning lines has been described, but television signals of other scanning systems (for example, PAL system, etc.) can be similarly processed. Furthermore, although the embodiment has been explained using a monochromatic television signal, 3J7? It is clear that similar conversion can be performed by providing similar circuits for each color signal, luminance, and color difference signal.

Further, the coefficients of the interpolation filter described in the embodiment are merely an example, and it is obvious that the present invention can be similarly implemented using other coefficients.

According to the present invention, in a portion of a television signal where a subject is stationary, a high-definition image with twice the number of scanning lines can be obtained by using information on all scanning lines of two fields spanning the frame. Furthermore, in the moving image part, the scanning lines within the field are used to double the number of scanning lines, resulting in images with less deterioration due to movement, converting current television signals into high-definition, high-quality images. It can be converted and displayed, and an economical high-definition television transmission system can be realized.

Furthermore, in the future, if high-definition televisions with twice the number of scanning lines from television cameras are realized, current television programs can be converted to high-definition television signals using the method of the present invention, so the number of programs that can be displayed on high-definition televisions will increase. , we can expect a great effect on the spread of high-definition television. If a television signal made high in definition by the processing method of the present invention is used for electronic printing, any television image can be printed as a high definition image with little deterioration.

[Brief explanation of the drawing]

Figure 1 is a block diagram showing the general configuration of a television receiver in which the scanning lines of the input television signal are doubled;
The figure is a block diagram for explaining the principle of the time axis conversion circuit 2 in Fig. 1 and a diagram showing the relationship of scanning lines in the field for explaining its operation. FIGS. 4 and 5 are diagrams showing the relationship between scanning lines in three fields of television signals; FIGS. 4 and 5 are circuit diagrams of an embodiment of means for obtaining an interpolation signal, which constitutes a main part of the television signal processing circuit according to the present invention; and FIG. , 7 and 8 are diagrams for explaining the operation of the embodiment of FIG. 5, respectively, and are diagrams showing the positional relationship between the scanning lines of the input television signal and the interpolated television signal, and the interpolation circuit. FIG. 6 is a diagram showing a frequency characteristic diagram of a filter to be formed and a relationship between scanning lines and coefficients of a coefficient circuit group in the embodiment of FIG. 5; DESCRIPTION OF SYMBOLS 1... Demodulation device, 2, 24.25... Time axis conversion circuit, 3... Synchronization signal conversion device, 4... Video amplification circuit, 5... Display device, 6.12... Input terminal, 7...Field memory, 8, 9, 24, 25...
- Time axis conversion circuit, 10.26... switching circuit, 11.
13... Output terminal, 14, 15, 16, 27, 28.
29...Delay memory, 17, 23, 36.37...
Addition circuit, 18.42... Coefficient circuit, 19, 38, 3
9... Subtraction circuit, 20... Coefficient circuit, 21, 22,
40.41... Multiplication circuit. Reference l Figure CA) (J)
((, Nki J Rone 4th V, S old

Claims (1)

[Claims] 1. In a television signal processing circuit which inputs an interlaced scanned television signal and obtains a high-speed scanned television signal by a scanning line interpolation circuit having a field memory and a time axis conversion circuit. A television signal processing circuit comprising a weighted addition of signals of a plurality of scanning lines constituting a frame image, and having a frequency characteristic in which a vertical high frequency component gain is larger than a DC gain. 2. The television signal processing circuit according to item 1, further comprising a motion detection circuit that detects the movement of a subject from the input television signal, and the scanning line interpolation circuit has a gain greater than a DC gain based on the output of the motion detection circuit. A television signal processing circuit characterized in that it is configured to change the gain of a vertical high frequency component according to motion.