JPH10161590A - Motion detecting circuit - Google Patents

Abstract

PROBLEM TO BE SOLVED: To inhibit the detection of a wrong motion vector caused by the inclusion of noise when the level of an input video signal is lower than a setting value. SOLUTION: A level conversion circuit 30 is coupled with the precedent stage of a motion detecting circuit 10 which detects a motion vector by the block matching method. When the level X of a video signal inputted to an input terminal 14 is less than a preset value S, a signal A(>1) of (A+1: e.g. 10)X is outputted to the detecting circuit 10 and when the level X is lower than S, the signal of (AS+X) is outputted to the detecting circuit 10. In the former case, the level X of each dot inside the block UB for determining correlation values is multiplied by (A+1) and the total sum of the level is also multiplied by (A+1). Therefore when noise mingles into the dots inside the block UB, the proportion of noise occupied in the total sum is lowered and the dots are difficult to be susceptible to the effects of noise to allow the high-precision detection of motion vectors.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a motion detecting circuit for detecting a motion vector used for moving image correction and the like. The motion vector detected by the motion detection circuit is displayed on a display device using a PDP (plasma display panel) or an LCD (liquid crystal display panel) (for example, a display device of an address / display separation type driving system). It is used for moving image correction to reduce a visual display shift that sometimes occurs.

[0002]

2. Description of the Related Art Recently, as a thin and lightweight display device, P
DP and LCD are attracting attention. The driving method of this PDP is completely different from the conventional CRT driving method, and is a direct driving method using a digitized input video signal. Therefore, the luminance gradation emitted from the panel surface is
It is determined by the number of bits of the signal to be handled.

[0003] PDPs are AC type and D type having different basic characteristics.
It is divided into two types of C type. In the AC type PDP, sufficient characteristics have been obtained with respect to luminance and life, but as for gradation display, only a maximum of 64 gradation display has been reported at the prototype level, but the address / display separation type driving method (ADS) A sub-field method) has been proposed. The drive sequence and drive waveform of the PDP used in this method are as shown in FIGS. 4A and 4B, for example.

FIG. 4A shows a drive sequence of 8 bits and 256 gradations. One frame has eight luminance ratios of 1, 2, 4, 8, 16, 32, 64 and 128. Subfields SF1, SF2, SF3, SF4, SF
5, SF6, SF7, and SF8, and a display of 256 gradations is performed by a combination of luminances of eight screens.

[0005] Each subfield is shown in FIG.
As shown in (1), an address period in which data for one refreshed screen is written and a sustain period for determining the luminance level of the subfield are constituted. In the address period, first, wall charges are initially formed on each pixel at the same time for the entire screen, and then a sustain pulse is applied to the entire screen to perform display. The brightness of the subfield is proportional to the number of sustain pulses and is set to a predetermined brightness. In this way, 256 gradation display is realized.

When a moving image is displayed on the display device of the address / display separation type driving system as described above, a visual display shift may occur. There is a moving image correction process using a vector. The motion detection circuit 10 for detecting such a motion vector has conventionally been configured as shown in FIG.

That is, in the motion detection circuit 10 shown in FIG. 5, the current frame screen based on the video signal (video data) input to the input terminal 14 by the frame memory 12
A video signal of a screen before a frame (hereinafter, referred to as a previous frame screen) is created, and the correlation value calculation unit 16 uses the unit block UB (for example, 8 ×) to be detected in the current frame screen.
8 dots), the correlation values (difference values) with all the unit blocks UB in the motion vector detection area of the previous frame screen are sequentially obtained. Generates a signal (hereinafter, simply referred to as a movement vector) indicating a moving direction and a moving amount from the unit block position to the detection target unit block position,
The movement vector is output from the output terminal 20 as the motion vector of the unit block to be detected.

[0008]

However, in the conventional example shown in FIG. 5, since the video signal is directly input to the motion detection circuit 10, when the input video signal level (hereinafter simply referred to as input level) is low. In addition, there is a problem that a detection error occurs due to the contamination of noise. That is, the sum of the block levels (the level of the unit block UB) is small in the screen portion where the input level is low, and the ratio of the mixed noise to the sum of the block levels increases.
The sum of the block levels changes significantly due to noise, and when a motion vector is detected based on the correlation value between the block levels of the current frame screen and the previous frame screen, a motion vector far from the true value is detected. There was a problem that it would.

For example, in the case where the maximum input level is 255 and the unit block UB is 3 × 3 dots, the true sum of the block levels is shown in FIG. It becomes 9 as shown. When noise of level 9 is mixed into one dot in such a unit block UB, the sum of the block levels becomes as shown in FIG.
As shown in (b), the noise ratio is 18 and the ratio of the noise to the total sum is as large as 50%. Therefore, there is a problem that an erroneous motion vector is detected due to the influence of the noise.

[0010] The present invention has been made in view of the above-described problems, and has been made in consideration of the above-described problems when the level of an input video signal is lower than a set value.
It is an object of the present invention to provide a motion detection circuit that does not detect an erroneous motion vector due to noise.

[0011]

According to the present invention, there is provided a motion detecting circuit which obtains a block level correlation value between a current frame screen and a previous frame screen based on an input video signal, and detects a motion vector based on the correlation value. In the above, a level conversion circuit is connected to the preceding stage of the motion detection circuit, and the level conversion circuit multiplies the input level X by a coefficient A of 1 or more when the level X of the input video signal is equal to or less than the set value S. A signal having a level (A + 1) X obtained by adding AX to the input level X is output. When the input level X is equal to or more than the set value S, a level (AS +
X) is output.

The level conversion circuit comprises, for example, an input level addition value setting section (for example, a lookup table) and an addition section. When the level (input level) X of the input video signal is lower than the set value S, the input level X Is multiplied by one or more coefficients A and a level (A
+1) The signal of X is output to the motion detection circuit. For this reason,
The level X of each dot in the unit block for obtaining the correlation value is (A + 1) times and input to the motion detection circuit, and the sum of the levels of the unit block UB (for example, 3 × 3 dots) is also (A + 1) times. Therefore, when noise is mixed in the dots in the unit block UB, the ratio of the noise to the total sum is reduced to make it less susceptible to the noise, and a highly accurate motion vector can be detected.

[0013]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of a motion detection circuit according to the present invention will be described below with reference to the drawings. In FIG. 1, the same parts as those in FIG. In FIG. 1, 10
Is a motion detection circuit. The motion detection circuit 10 has the same configuration as that of FIG. A level conversion circuit 30 unique to the present invention is provided at a stage preceding the motion detection circuit 10.

The level conversion circuit 30 comprises an input level addition value setting section 32 and an addition section 34, and is configured to have a conversion characteristic as shown by a dotted line P in FIG. That is, when the level (hereinafter simply referred to as input level) X of the video signal input to the input terminal 14 is equal to or less than the set value S, the level AX obtained by multiplying the input level X by a coefficient A of 1 or more
Is added to the input level X, a signal of the level (A + 1) X is output (Y = (A + 1) X), and the input level X is set to the set value S
At this time, a signal of a level (AS + X) obtained by adding the constant level AS to the input level X is output (Y = AS + X).

The input level addition value setting section 32 is constituted by a look-up table (LUT) that uses the level X of the video signal input to the input terminal 14 as an address and outputs the contents of the address, and outputs the result (Y ) Has input / output characteristics as shown by the solid line Q in FIG. That is,
When the input level X is equal to or less than the set value S, the input level X is set to 1
The signal of the level AX multiplied by the coefficient A is output (Y = A
X), and outputs a signal of a constant level AS (Y = AS) when the input level X is equal to or higher than the set value S.

The adder 34 is configured to add an input video signal to an output signal of the input level addition value setting unit 32 and output the converted signal to the motion detection circuit 10 as a converted output. For this reason, a signal that satisfies the conversion characteristics indicated by the dotted line P in FIG. 2 is output from the output side of the adder 34. In FIG. 2, R indicated by a two-dot chain line indicates a reference characteristic indicating a relationship of output = input (Y = X).

Next, the operation of FIG. 1 will be described with reference to FIGS.
Will be described together. For convenience of explanation, unit block UB
Has a 3 × 3 dot configuration, and the maximum value of the input level X is 25.
When the set value S at 5 (FF in hexadecimal notation) is 32 (16
20), the coefficient A is 9, and the input level X of the video signal input to the input terminal 14 is 1.

In this case, the input level X of the video signal input to the input terminal 14 is 9
Since it is multiplied (A = 9) and input to one input side of the adding unit 34, it is multiplied by 10 (= A +
The signal of level 10 of 1) is output and input to the motion detection circuit 10. For this reason, the level corresponding to each dot in the unit block UB becomes 10, as shown in FIG.
The sum of the levels is 100.

If it is assumed that noise of level 10 is mixed in an arbitrary dot in the unit block UB, the dot level becomes 20 as shown by hatching in FIG. Level sum is 1
00, the ratio of noise to the total is 1
This makes it possible to detect a highly accurate motion vector with less influence of noise.

In the above embodiment, the case where the input level addition value setting section is constituted by a look-up table in which the input level X is an address and the contents of the address is an output is described, but the present invention is not limited to this. Not a thing,
When the input level X is equal to or less than the set value S, the input level X is set to 1
A configuration in which a signal of the level AX multiplied by the coefficient A is output and a signal of the constant level AS is output when the input level X is equal to or higher than the set value S (for example, a device mainly composed of a ROM or a RAM) Can also be used.

In the above embodiment, the case where the level conversion circuit is constituted by the input level addition value setting section and the addition section has been described. However, the present invention is not limited to this, and the input level X is equal to or less than the setting value S. When the input level X is higher than the set value S, a signal of a level (A + 1) X obtained by adding a level AX obtained by multiplying the input level X by a coefficient A of 1 or more to the input level X is output. What is necessary is just to output a signal of the level (AS + X) obtained by adding AS to the input level X.

[0022]

According to the present invention, a level conversion circuit is connected in front of a motion detection circuit for detecting a motion vector by the block matching method. This level conversion circuit is used when the level X of the input video signal is equal to or less than a set value S. And outputs a signal of level (A + 1) X obtained by adding a level AX obtained by multiplying the input level X by a coefficient A of 1 or more to the input level X. When the input level X is equal to or higher than the set value S, a constant level AS is output. Since the signal of the level (AS + X) added to the input level X is output, when the level X of the input video signal is lower than the set value S, the input level X is multiplied by (A + 1) (for example, 10 times).
This signal can be output to the motion detection circuit.

For this reason, the level X of each dot in the unit block for obtaining the correlation value is also multiplied by (A + 1) and input to the motion detection circuit, and the unit block UB (for example, 3 × 3)
Since the sum of the levels of (dots) is also (A + 1) times, when noise is mixed in the dots in the unit block UB, the ratio of the noise to the total is reduced to make the noise less susceptible to the noise, and the motion with high accuracy Vectors can be detected.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an embodiment of a motion detection circuit according to the present invention.

FIG. 2 is a characteristic diagram illustrating the operation of FIG.

FIG. 3 is a diagram showing a signal level that is multiplied by 10 (when A = 9) and output to the motion detection circuit 10 by the level conversion circuit 30 when a video signal whose input level X is 1 is input to the input terminal 14 of FIG. (A) is an explanatory diagram of the level of each dot in the unit block UB of the previous frame screen before noise mixing, and (b) is a dot explanatory diagram of the corresponding unit block UB of the current frame screen after noise mixing. FIG.

FIG. 4 is a diagram for explaining a subfield lighting method;
(A) is an explanatory diagram of a drive sequence in a 256-tone method, and (b) is a drive waveform diagram.

FIG. 5 is a block diagram of a conventional motion detection circuit.

6 is an explanatory diagram of signal levels when a video signal whose input level X is 1 is directly input to the motion detection circuit 10 of FIG. 5;
(A) is a level explanatory diagram of each dot in the unit block UB of the previous frame screen before noise mixing, and (b) is a level explanatory diagram of each dot in the corresponding unit block UB of the current frame screen after noise mixing. is there.

[Explanation of symbols]

10: motion detection circuit, 12: frame memory, 14
... video signal input terminal, 16 ... correlation value calculator, 18
... Motion vector generation unit, 20 ... Motion vector output terminal, 30 ... Level conversion circuit, 32 ... Input level addition value setting unit, 34 ... Addition unit, A ... One or more coefficients,
P, Q, R: characteristic line, S: set value, UB: unit block, X: input level, Y: output level.

Claims (3)

[Claims] 1. A motion detection circuit which calculates a block level correlation value between a current frame screen and a previous frame screen based on an input video signal and detects a motion vector based on the correlation value. When the level X of the input video signal is equal to or less than a set value S, the level conversion circuit converts the level AX obtained by multiplying the input level X by a coefficient A of 1 or more into the input level X. And outputs a signal of level (A + 1) X, which is added to the fixed level A when the input level X is equal to or higher than the set value S.
A motion detection circuit which outputs a signal of a level (AS + X) obtained by adding S to the input level X. 2. A level conversion circuit comprising an input level addition value setting section and an addition section, wherein the input level addition value setting section includes a coefficient equal to or more than 1 when the input level X is equal to or less than a set value S. A, and outputs a signal of a level AX. When the input level X is equal to or higher than a set value S, a constant level AS is output.
2. The motion detection circuit according to claim 1, wherein the addition section outputs an output signal of the input level addition value setting section and an input video signal to generate a converted output. 3. An input level addition value setting section, comprising:
3. The motion detecting circuit according to claim 2, wherein the address is used as an address, and the content of the address is used as a look-up table for output.