JPS62230277A - Scanning line converting device - Google Patents

Abstract

PURPOSE:To reduce the deterioration of a picture in the vicinity of the boundary of a still picture and an animation by controlling the mixing ratio between the signal-group of current filed and those of the fields immediately before/after the current field by means of a motion decision signal from a motion detection circuit. CONSTITUTION:A video signal having 1125 pieces of scanning lines and inter laced by ratio 2:1 is supplied to the first stages of field memories 150 and 151 connected in series through in input terminal 100, and at the same time supplied to a motion detecting circuit 160 and a timing signal generating circuit 200. An output from the memory 151 is added with a signal S102 in an adding circuit 152, then halved. The signal S of the current field is supplied to line memories L21-L29. The outputs of the respective memories L21-L29 are averaged, then transferred to mixers 51-59. In the mixers 51-59, the signals of current field and those of the fields immediataly before/after the current field, and the ratio or mixing is controlled by means of a motion decision signal from the motion detecting circuit 160.

Description

[Detailed Description of the Invention] [Objective of the Invention] (Industrial Application Field) This invention provides a so-called high-definition television that has more scanning lines than the current 525 (NTSC system) or 625 (PAL system). For devices with fewer scanning lines than the signal, for example, 525 or 625 scanning lines.

The present invention relates to a scanning line conversion device used when receiving an image with a device such as the above.

(Prior art) Conventionally, when high-definition television broadcasting began to be carried out using broadcasting satellites, it was the same as in the early days when black-and-white television broadcasting and color television broadcasting were time-divided and broadcast on the same channel. As such, it is expected that current system broadcasting and high-definition television broadcasting will be performed in a time-sharing manner. Also, at the beginning of high-definition television broadcasting,
It can be predicted that the penetration rate of high-definition television receivers will also be low. Therefore, there is a desire to view high definition television broadcasts using conventional, current type television receivers.

However, high-definition television signals only have 1125 lines, whereas conventional signals only have 5 lines.
25 or 625 lines, so in order to watch high-definition television broadcasts on a conventional television receiver, it is necessary to perform scan line conversion while preserving the video quality as much as possible in terms of resolution and naturalness of moving images. There is.

The following two conventional methods of converting the number of scanning lines are known:
There is one.

■ A method of converting running lines within the field.

(2) For example, when using a 2:1 interlaced scanning method, image information for two fields is first converted into sequential scanning, and then processing within the frame is performed.

Here, in the method (2), the resolution in the vertical direction deteriorates, especially when the conversion ratio of the number of scanning lines is 2=1 or more. Furthermore, in the case of method (2), since it is necessary to first convert to sequential scanning, there is a problem that the hardware scale becomes large.

In order to solve the above problems,
There is a scanning line number conversion method described in Japanese Patent No. 6582.

Conventionally, as shown in FIG. 7, in this method, a progressive scan conversion circuit 71 converts a signal into a sequential scan signal, passes it through an interpolation filter 72 and a thinning filter 73, and then converts the signal into an interlaced scan signal again in an interlace conversion circuit 74. This further improves the configuration of converting into . That is, as shown in FIG. 9, scanning line interpolation and thinning scanning are performed at the same time, eliminating the need for a sequential scanning conversion circuit and minimizing resolution degradation in the vertical direction. In this way,
In contrast to the conversion circuit shown in FIG. 7, where each block is realized by hardware and the scale of multiple circuits becomes large, the scale of the hardware can be reduced.

(Problem that the invention attempts to solve) This will be explained with reference to the scanning line conversion circuit shown in FIG.

When signal 8801 is the signal of one field later, signal 5802 is the signal of the current field, and signal 5803 is the signal of one field before, signal 5804 is the signal of signals 5801, 88
This is a signal that is multiplied by 0.03.

To create signals 5802 and 8803, field memo 1
) 81.82 is used and adder 8 is used to create 5804.
3 is used.

The signal 5804 is a plurality of line memos 1) L connected in series.
The signal 5802 is supplied to the first stages of line memories L21 to L29 connected in series.

The input/output side signals X1 to xn, y, to Yn of each line memory Lll to L19 and L21 to L29 are shown in FIG.
It is supplied to each coefficient adder 85-89. Then, the output of each coefficient adder 85 to 89 is selected by a switch 90, and
The output signal is scan line converted.

In the above system, to perform scan line interpolation,
Since scan line interpolation is performed not only by using the signal 8802 of the current field but also by using the signal 8804 of a different field, degradation in vertical resolution can be minimized.

However, considering that television images include still images and moving images, the configuration shown in FIG. 8 allows scan line conversion with minimal vertical resolution degradation; however, in the case of moving images, Because the signal 5804 of a different field is used in addition to the signal 5802 of the current field, the moving images overlap, resulting in, for example, two-line blurring, and contrary to the above expectations, the resolution of a moving image is rather reduced. There is. Therefore, if a 6-scan line conversion device that processes still images and moving images separately is realized, the scale of the hardware will become large and it is not practical.

Furthermore, other problems when processing moving images will be explained. For simplicity, consider converting the number of scanning lines from four to three.

Figure 11 shows the number of scanning lines: 4 ((a) in the same figure) and 3 (((a) in the same figure).
b) shows the vertical frequency spectrum of). In order to perform scanning line conversion without turning back, it is desirable to use a filter with steep frequency characteristics, such as having a pass band in the shaded area in the figure. However, in general, implementing a filter with such characteristics requires a very large number of taps. The conventional coefficient adding circuits 85 to 89 (filters)
As an example is shown in Figure 0, kXl, kyl, -・
- Has taps of kXn and kYn.

Next, in order to explain the filter action, it will be explained with reference to FIG.

FIG. 12 shows sample points in the vertical direction one field before, the current field, and one field after. The circles in the figure indicate sample points on one scanning line, and it is assumed that the black part can move upward.

In the figure, if a filtering effect is applied to the data of sample group ■, the current field data is a % f, and the data one field before is k f
% Of, data after 1 field is kb ~ O
b will be used. Here, the explanation will be given assuming that still image processing and video processing are performed separately depending on the movement of the image. If motion detection is to be performed between samples of the pixel (2) of interest, a difference calculation is performed between each frame, and in the illustrated example, mb - mf does not become zero. It is determined that the video is a temporary video. For video, scanline conversion is performed using samples within the current field. As a result, between samples a'' and c, it is originally necessary to also use kf, Jf, kb, /b, but since interpolation processing is performed only with data from a to c, the resolution in the vertical direction is degraded. You will end up letting it happen.

Next, we move on to filtering sample group ■.
Consider another case. In this case, when motion detection is performed between samples of the pixel of interest l, n f −n b =Q. Therefore, in this case, it is determined that the image is a still image, and interpolation processing using Ijf-pf and 7b-pb is performed. For this reason, a two-line blur will occur on the screen near the black circle.

As mentioned above, in order to convert the number of scanning lines while preserving the quality of the video signal such as resolution and naturalness of moving parts, it is necessary to perform video processing for moving parts and still image processing for still parts. However, the configuration shown in FIG. 8 has a problem in that it becomes less suitable as the number of taps increases.

Therefore, it is an object of the present invention to provide a scanning line conversion device that performs scanning line number conversion processing in accordance with still images and moving images, and in particular does not cause two-line blurring for moving images.

A further object of the present invention is to provide a scanning line conversion device that does not have an adverse effect on images around moving parts after conversion due to an increase in the number of taps of a digital filter used for scanning line conversion.

[Structure of the invention]

(Means for Solving the Problems) In the present invention, for example, as shown in FIG. and a signal group 513n created using the previous and next scanning lines in the current field.
The signal group Xn is created by controlling the mixing ratio of the signal groups 510n and 513n using the motion determination signal group 512n from the motion detection circuit. Then, the signal groups Yn and Xn are supplied to a coefficient adder circuit for weighting and addition, and a corresponding addition output (scanning line number conversion output) is extracted at a relative vertical position according to a timing signal. It is something.

(Function) With the above means, the tap coefficients of the digital filter used for scanning line conversion can be used in common for still images and moving images. In addition, it is possible to perform motion detection for each scanning line and adjust the mixing ratio of appropriate line or field interpolated signals or interpolated signals for each scanning line, thereby obtaining a good image. I can do it.

(Example) Examples of the present invention will be described below with reference to the drawings. 1st
The figure shows an example of the present invention.
This will be explained as conversion to a 1:1 interlaced video signal. Input terminal 100, number of scanning lines 1125, interlace ratio 2
:1 video signal is supplied. The signal at this input terminal 100 is transmitted to field memories 150 and 151 connected in series.
It is also supplied to the motion detection circuit 160 and the timing signal generation circuit 200 that generates the switching timing signal. The output of field memory 151 is added to signal 5102 in adder 152 and multiplied by 7. The output signal 5105 of this adder 152 is input to the first stage of a plurality of line memories Lll-L19 connected in series, and is used as an odd field signal. Further, the current field signal 8103 is input to the first stage of a plurality of line memories L21 to L29 connected in series, and is used as an even field signal.

Now, the signals of the input/output section of each memory are set to 810 as shown in the figure.
2 to 8109.8110-4119. Then, if the signal 5103 is the signal of the current field, the signal SJ 04 of one field before, and the signal 5102 after one field.
is also supplied to the motion detection circuit 160.

The motion detection circuit 16θ performs motion detection by a method described, for example, in a document (Television Society Technical Report TEB883-4, Study of Interpolation Filter for Interlaced Progressive Scan Conversion).

Motion determination signals 5121 to 51 from the motion detection circuit 16θ
29 are supplied to the control ends of the corresponding mixers 51 to 59, respectively. At one input end of each of the mixers 51 to 59,
Signals 5105 to 5109 are respectively supplied, and signals 5131 to 5139 are respectively supplied to the other input terminal. Signals 5131 to 5139 are input to the input side and output side of the line memories L21 to L29, respectively, and added to the adder 15.
3 to 159 and multiplied by one.

Next, we will discuss signal formation operations during still image processing and video processing.
I will explain.

In the case of a still image, the even field signals supplied to line memo +) L2'l to L29 are a 0 horizontal scanning period delayed signal (hereinafter referred to as a QH delayed signal) Yl, and a 1 horizontal scanning period delayed signal (hereinafter referred to as an IH delayed signal). That's what it means.

Same below)Y2. . . . is derived as a (n-+)H delayed signal Yn and supplied to the coefficient adder circuit of FIG.

Alternatively, the odd field signal 5105 is sent to the line memories L11 to L19 as an OH delay signal 5105, an IH delay signal 5106, and -(nt)H delay signal 8109.
from the group and supplied to mixers 51-59.

Here, if the motion determination signal 8121 that is the output of the motion detection circuit 160 is a signal indicating that it is a completely still image,
In mixer 51, signal 5105 is selected and derived as signal X. Similarly, the selection of each mixer 52-59 is determined by motion determination signals 5122-5129, which are derived as signals x2-Xn and supplied to the coefficient adder circuit (shown in FIG. 2).

Further, motion determination signals 5121 to 5 of the motion detection circuit 160
If 129 indicates a moving image, the mixer 5
1 to 59 select signals 5131 to 8139 and signal X1
~Xn.

FIG. 2 shows the coefficient adders 21 to 29, and FIG.
For example, one of the coefficient adders 21 is shown as a representative. That is, even field signals Y1 to Yn and odd field signals X1 to Xn are processed by coefficient multipliers M1 to M2n, respectively.
After passing through the adder A, the signals are combined and output. FIG. 4 further shows the coefficient multiplication process by extracting some lines, and shows the signal Y1 based on the even field signal 5103.
~Y, is multiplied by the coefficient E1~E, and the signal X, ~X4 based on the odd field signal 5105 is multiplied by the coefficient θ□~04
is multiplied by

The coefficient adder circuit outputs X1 to Xn from each line memory so that the conversion from one scanning line to m scanning lines can be performed without folding back.

Weighting is performed on Y1 to Yn. The switching device 31 shown in FIG. 2 is controlled by the timing signal 5200 generated by the switching timing signal generation circuit 200 shown in FIG. The switch 31 selects the output of each coefficient adder and obtains output signals of m scanning lines.

In the case of video processing, the following conversion processing is performed. This is scanning line conversion using only signals of the same field in order to prevent two-line blurring. The even field signal 5103 shown in FIG.
~27.

In this embodiment, since the conversion is from a 2:1 interlaced video signal with 1125 scanning lines to a 2:1 interlaced video signal with 525 scanning lines, the conversion ratio is 15 near. Therefore, while the former scans 15 lines, the latter scans 7 lines, so in this embodiment, 7 phases are required as data after conversion. Therefore, the coefficient adders shown in FIG. 2 21 to 27 are required to generate data having each of these seven phases. Furthermore, in the case of a moving image, the signal 51 in FIG.
31-5139 are taken out via mixers 51-59 and used as signals X, -Xn. In other words, only signals of the same field are used, and signals of each line and interpolated signals between lines are used.

The calculation form indicated by the broken line in FIG.
8139 is shown. That is, signals Y1 to Y correspond to even fields in a still image.

is similarly subjected to multiplication processing in the coefficient multipliers E1 to E in the moving image. However, in the case of a still image, the coefficient multipliers 01 to 04 that processed the odd field signals have signals 5131 to 513 interpolated using the even field signals Y, ~Y.
9 are supplied as signals X, .about.Xn. In the subsequent processing, as in the case of the still image described above, switching is performed by the switch 31, and a 525-line video signal with an interlace ratio of 2:1 is created.

As described above, in the present invention, the tap coefficients of the filter for scanning line conversion can be shared between still image processing and moving image processing. Therefore, there is no need to expand the hardware.

Next, operations related to the second object of the present invention will be described. In this invention, the motion detection circuit 160 obtains motion determination signals 5121 to 5129 for the signals S1θ5 to 5109 for each line, and
He is trying to control the mixing ratio of each mixer 51-59 according to the size of 21-5129. Therefore, even within the same field, appropriate scanning line signals can be obtained depending on the moving image portion and the still image portion, and blurring between two lines will not occur. This operation will be explained with reference to FIG. Regardless of whether the even field signal 5103 is a still image or a moving image, signals Y1 to Y4 are multiplied by coefficients in coefficient multipliers E1 to E4. On the other hand, regarding the odd field signal 5105, motion detection is performed for each of the signals X and X4, and if the signal X is determined to be a still image S, this signal X1 is directly supplied to the coefficient unit 01. Ru. However, when it is determined that it is a moving image M, as in the line of signal X2, even field signal Y2. A signal X2 created using Y3 is supplied to a coefficient unit 02. Similarly, the signal
When 3 is a moving image, the signal x3 is supplied to the coefficient unit 03,
When the signal X4 is a still image, the signal X4 becomes coefficient 0 as it is.
4.

FIG. 6 shows yet another embodiment of the invention. The same parts as in FIG. 1 will be described with the same reference numerals.

In this embodiment, an interpolation signal between lines is generated in advance using a line memo+) L910 and an adder 900, and this is supplied to the first stage of line memories L911 to L919. As a result, the signals 5131-51
39 is always ready. In addition, line memory L92
0-L920 is used as a dedicated memory for creating signals Y, -Yn for even fields.

〔Effect of the invention〕

As explained above, the present invention provides scan line conversion that can reduce image deterioration near the boundary between still images and moving images, while minimizing deterioration in vertical resolution of video signals and without increasing hardware. equipment can be provided.

[Brief explanation of drawings]

FIG. 1 is a circuit diagram showing an embodiment of the present invention, FIG. 2 is a diagram showing a coefficient adder circuit used in the circuit of the present invention,
FIG. 3 is a diagram showing one of the coefficient adders shown in FIG. Fig. 6 is a circuit diagram showing another embodiment of the present invention, and Figs. A circuit configuration diagram shown to explain the scanning line conversion device, FIG. 11 is an explanatory diagram of the scanning line conversion principle,
FIG. 12 is an explanatory diagram of the principle of still image/moving image determination processing. 51-59...Mixer, 150.151...Field memory, 152-159...Adder, 160...
-Motion detection circuit, 200...Switching timing signal generation circuit, L11 to L19. L21-L29...Line memory, 21-29...Coefficient adder, 3...Switcher Applicant's agent Patent attorney Takehiko Suzue Figure 3 Signal 5t03 Figure 5 Figure 10 Amplitude ( b) Conversion from 3 scanning lines to 4 scanning lines

Claims (1)

[Scope of Claim] An apparatus for converting the number of scan lines of a television signal having a greater number of scan lines than an interlaced current format television signal into the number of scan lines of the current format television signal, comprising: scanning within a current field; means for obtaining in parallel a signal group Yn of a line group; means for obtaining in parallel a signal group S10n obtained by adding scanning lines at the same position in the vertical direction using at least scanning line groups in fields before and after the current field; Signal group S1 interpolated by line interpolation using signal group Yn so that it is at the same position in the vertical direction as the scanning line group of the previous and subsequent fields.
3n in parallel; means for obtaining a motion determination signal group S12n that detects the movement of an image due to a video signal for each scanning line using at least a scanning line group between fields before and after the current field; Group S10n, signal group S11n, signal group S13n
and motion determination signal group S12n are respectively made to correspond in the vertical direction, the signal group S10n and the signal group S13n are supplied to a mixing input section, the mixing ratio is determined according to the content of the motion determination signal group S12n, and the signal group is outputted. means for obtaining Xn, weighting and adding the signal groups Yn and Xn in a coefficient adder, and vertically converting each addition output so that the number of scanning lines before scan conversion becomes a desired number of scanning lines; 1. A scanning line conversion device comprising: means for switching and selecting and deriving data according to relative positions.