JPH07112273B2 - Band compression transmitter - Google Patents

Description

The present invention relates to a band compression transmission device applied to transmission of high definition television signals.

B. Outline of the Invention The present invention calculates the motion vector of this motion and transmits the motion vector when the entire screen moves, such as when a pixel is thinned out and transmitted for each field and the television camera pans. On the receiving side, however, it is a band compression transmission device that performs motion compensation based on the motion vector, and the resolution of the motion vector is the clock cycle of the sampling frequency 2fs, and the resolution of the motion compensation is the clock of the sub-sampling frequency fs. In the case of a device that controls the delay circuit for converting the calculated motion vector into a motion correction amount with a resolution of 1 / fs in the case of a cycle, the control of the above delay circuit provided on the transmission side and the reception side, respectively. That is, when controlling whether the delay circuit is inserted or not inserted in the data transmission path, the delay is delayed when a still image is obtained by observing the motion vectors of two consecutive fields. By prohibiting inverted and is obtained so as to reduce the influence of malfunction of the calculation of the motion vector due to noise.

Further, according to the present invention, the delay circuits provided on the transmitting side and the receiving side are also used for the television signals of the odd field and the even field, thereby further simplifying the configuration.

C. Conventional technology A high-definition television has 1125 horizontal scanning lines, which is about twice the number of horizontal scanning lines, and the bandwidth of the video signal reaches 20 MHz for the luminance signal. Therefore, a system (called a MUSE system) capable of transmitting a high-definition television signal with only one channel of satellite broadcasting has been proposed. According to this band compression transmission method, a high-definition television signal can be compressed to about 8 MHz.

In the MUSE method, the screen of four consecutive fields is changed for each field, and the pixels are thinned to 1/4 for transmission. That is, FIG. 7 shows a pixel thinning process. In FIG. 7, a pixel indicated by a circle indicates a pixel transmitted in the first field, and a pixel indicated by a triangle indicates a pixel transmitted in the second field. , ● indicates a pixel transmitted in the third field, and ▲ indicates a pixel transmitted in the fourth field. These pixels are in each field
It was thinned out to 1/4. X indicates pixels that are not transmitted, and these pixels that are not transmitted are reproduced on the receiving side by temporal and spatial interpolation with motion compensation and motion detection.

That is, in the MUSE method, in a moving part, an image looks unnatural due to the superposition of images of four fields.
In the moving area, the image is restored by using only the data for one field sent at that time.

Further, in the case of a movement of the entire screen at a constant speed due to the panning of the television camera or the like, a motion vector is obtained by the block matching method from the difference between the data of two consecutive frames, and this motion vector is used as the transmission signal. It is superimposed as control data. Of all 32 bits of control data, 5 bits from 1 to 5 are horizontal motion vectors, and 3 bits from 6 to 8 are vertical motion vectors. On the receiving side, motion compensation is performed to control the address for reading the received data from the memory according to this motion vector.

As shown in FIG. 8, a table of absolute value frame difference integration data between two consecutive frames is formed in a range of 8 lines in the vertical direction centering on the origin and 32 pixels in the horizontal direction,
The minimum value is calculated from this range, and its position (x, y)
Is detected as a motion vector.

High definition television signals are digitized with a sampling frequency of 2fs (64.8MHz). Therefore, the motion detection is also performed with the resolution of the pixel digitized at the sampling frequency of 2fs. However, the resolution of motion compensation on the receiving side corresponds to the sub-sampling rate of fs (32.4 MHz).

The resolution of motion vector detection is 1 / (2fs), whereas the resolution of motion correction is 1 / fs, so it is necessary to convert the detected motion vector to the motion correction amount of 1 / fs resolution. is there. This conversion can be performed by inserting or not inserting a delay circuit with a delay amount of 1 / (2fs) in the data transmission path.

FIG. 9A is an enlarged view of the scanning start portion (upper left of the screen) of the high-definition television screen, and the pixel interval is 2 fs (64 MHz
It corresponds to the sampling frequency of z). Pixels in upper case indicate pixels to be transmitted, pixels in lower case are not transmitted, and pixels formed by interpolation are shown. In the first field, the pixels marked with a circle are transmitted. Since the motion vector is detected based on the difference between the first field and the third field and the difference between the second field and the fourth field, FIG. 9 shows the motion between the first field and the third field, for example. Shows. In addition, a process of switching between a digital television signal that has passed through the one-clock delay circuit and a digital television signal that does not pass through it by the selection circuit is performed. Generally ON to OF
Switching from F or OFF to ON is called inversion, so when turning the state through the delay circuit ON and turning the state not through the delay circuit OFF, ON → OFF or OFF → O
The change in the state of N will be referred to as delay inversion in the following description.

Assuming that FIG. 9A is the screen of the first field that has not been subjected to the delay inversion process, for example, when the leftward movement of 4 clock cycles of the sampling clock of 2fs is in the third field, the screen of the third field is displayed when it occurs. Is as shown in FIG. 9B. In this case, since the motion is an integral multiple of 1 / fs, it is not necessary to perform delay processing by the 1-clock delay circuit. The sub-sampling phase also matches the format shown in FIG. 7, and the sub-sampling phase may be regularly inverted.

When movement of the sampling clock of 1 / (2fs) to the left by two clocks occurs in the third field, the screen shown in FIG. 9C is displayed. Since this movement is an integral multiple of 1 / fs, it is not necessary to perform delay processing by the 1-clock delay circuit. The sub-sampling phase is different from the format shown in FIG. 7, and the sub-sampling phase is the same as the previous frame.

When a movement of 1 / (2fs) to the left by one clock occurs in the third field, the screen shown in FIG. 9D is displayed. Since this movement cannot be expressed with the resolution of the clock of 32 MHz, the one-clock delay circuit is inserted and the delay processing is executed. In other words, due to the delay of 1 / (2fs),
The screen shown in FIG. D is converted to the screen shown in FIG. 9A, that is, the screen having no motion. The sub-sampling phase after the delay processing matches the format shown in FIG. 7, and the sub-sampling phase may be regularly inverted.

The relationship shown in FIG. 9 is the same for the second field and the fourth field. This invention is 1 /
It relates to the control of the delay circuit with the delay amount of (2fs).

D. Problems to be Solved by the Invention In the conventional band compression device, since the motion vector is obtained between consecutive odd fields and even fields separated by one frame time,
An independent 1-clock delay circuit was provided and controlled separately.

In addition, when the detection of the motion vector is wrong due to noise etc.,
Delay is not even in the odd and even fields,
There was a drawback that a phase shift of 1 / (2fs) remained between the two.

For example, in reality, even if it is a still image, it is erroneously detected by noise as if there was a movement of 1 / (2fs) in the third field, and there is no movement in the fourth and subsequent fields. If it is detected, a 1-clock delay circuit for data delay of the odd field is inserted in the third field, and this state continues in the fifth and subsequent odd fields. However, since the delay circuit is not inserted for the even field data, a phase shift of 1 / (2fs) always exists between the odd field data and the even field data.

Therefore, an object of the present invention is to provide a delay circuit with a delay amount of 1 / (2fs) commonly for odd fields and even fields, and to prevent malfunction due to noise, perform delay inversion processing when there is no motion in the still image. It is an object of the present invention to provide a band compression transmission device which can be prohibited to prevent the data phase from shifting between the odd field and the even field.

E. Means for Solving the Problems The present invention relates to a band compression transmission apparatus for compressing and transmitting pixels of a digital television signal having a sampling frequency of 2f _s by sub-sampling, in a continuous frame of a digital television signal. The resolution is 1 / (2f
s) means for detecting a motion vector, delay means having a delay amount of 1 / (2fs) to which the digital television signal is supplied, and one of the output signal of the delay means and the digital television signal selectively. The selecting means to be taken out, the sub-sampling means to which the output signal of the selecting means is supplied, the least significant bit “0” and “1” of the motion vector are determined, and whether or not the state of the selecting means is changed to the previous field Is detected, the delay control means for generating a control signal for controlling the selecting means and the motion vector of two consecutive fields are checked to determine whether or not the image is a still image. A band compression transmission apparatus comprising means for prohibiting change and means for adding a motion vector and a control signal to the output signal of the sub-sampling means.

Further, the present invention is a band compression transmission apparatus characterized in that a delay circuit 9 is provided in common for an odd field and an even field of a digital television signal.

F. Action A delay control unit 4 is provided on the transmission side, the calculated horizontal motion vector is supplied to this delay control unit 4, and a motion vector having a resolution of 1 / (2fs) is set to a motion correction amount of 1 / fs and It is converted into a delay control signal for controlling the delay circuit 9 of / (2fs). This delay control signal is generated according to the least significant bits "0" and "1" of the motion vector and whether or not the delay was inverted last time. In this case, delay inversion is forcibly prohibited when a still image is obtained from the motion vector of two consecutive fields.

The motion compensation amount and the delay control signal are added to the transmitted digital television signal. On the receiving side, while performing motion compensation based on the motion compensation amount, the delay control signal
The delay circuit provided on the receiving side is controlled.

Both the transmitting side and the receiving side are provided with a common delay circuit for odd field data and even field data.

G. Embodiment G1. Configuration of Transmission Side FIG. 2 shows the overall configuration of the transmission side in which an embodiment of the present invention is provided. In FIG. 2, a high-definition television signal digitized at a sampling frequency of 2 fs (64.8 MHz) is supplied from an input terminal designated by 1.

The high-definition television signal is supplied to the frame memory 2 and written in the frame memory 2. The motion vector calculation circuit 3 calculates a motion vector for each field from the data stored in the frame memory 2. This motion vector has a resolution of 1 / (2fs), and has a range of (-16 to +15) pixels / frame in the horizontal direction and a range of (-4 to -3) lines / frame in the vertical direction. It is a thing.

The calculated motion vector is supplied to the delay control unit 4 described later. The delay control unit 4 forms a control signal for the switch circuit 5. The digital television signal from the input terminal 1 is supplied to one input terminal 6 of the switch circuit 5, and 1 / (2f is supplied to the other input terminal 7 of the switch circuit 5.
The digital television signal is supplied through the delay circuit 9 having the delay amount of s). However, although not shown, it is assumed that the signals sent to the input terminals 6 and 9 are delayed by the processing time in the motion vector calculation and delay control.

The digital television signal from the output terminal 8 of the switch circuit 5 is supplied to the prefilter / subsample circuit 10. The pre-filter / sub-sampling circuit 10 includes a pre-filter for each of the static region and the moving region, a mixer for switching the output of each pre-filter, and a sub-sampling circuit. As shown in FIG. 7 from the pre-filter / sub-sampling circuit 10, the data decimated to 1/4 is supplied to the adding circuit 11.

The motion vector formed by the motion vector calculation circuit 3 is
It has 5 bits in the horizontal direction and 3 bits in the vertical direction. The delay control section 4 forms a 1-bit control signal for controlling ON / OFF of the delay circuit 9 based on the least significant bit (LSB) of horizontal 5-bit data and the previous delay control.

The 5 bits of the horizontal motion vector are converted into a 4-bit motion correction amount using the delay control signal of the current field and the delay control signal of the previous field (the delay control signal is set to "1").
If the transmission side delay is defined as ON, horizontal correction (4 bits) = horizontal movement (5 bits) + current field delay signal (1 bit) -preceding field delay (1 bit) is obtained as the upper 4 bits. . ) The 1-bit delay control signal and the 3-bit vertical motion vector are combined by the adder circuit 12, and further added by the adder circuit 11 to the transmission data. The digital data transmitted to the output terminal 13 of the adder circuit 11 is taken out.

G2. Configuration of receiving side As mentioned above, band-compressed high-definition television signals are transmitted by satellite broadcasting. The decoder of the band compression device provided on the receiving side has the configuration shown in FIG.

In FIG. 3, the digital high-definition television signal received by the separation circuit 15 is supplied from the input terminal indicated by 14. The received data passed through the separation circuit 15 is supplied to the frame memory 16. The frame memory 16 includes a separation circuit 15
The amount of motion correction is supplied from and the motion correction is performed. Further, the frame memory 16 converts the received data of the 1/2 fs sub-sampling rate into the data of the fs sampling rate. The data of the sub sampling rate of fs is supplied to the interpolation / mixer circuit 17.

The interpolation / mixer circuit 17 interpolates data that was not transmitted by using continuous 4 fields of data in the static region, and used data of 1 field in the dynamic region that was not transmitted. It is composed of an interpolation filter for interpolating data and a mixer for switching these interpolation outputs. The output data of the interpolation / mixer circuit 17 has a sampling rate of 2fs.

The output data of the interpolation / mixer circuit 17 is supplied to one input terminal 19 of the switch circuit 18, and also to the other input terminal 20 of the switch circuit 18 via a delay circuit 22 having a delay amount of 1 / (2fs). Supplied. The output terminal 21 of the switch circuit 18 is derived as the output terminal 23. The switch circuit 18 is controlled by the 1-bit control signal from the separation circuit 15. The ON / OFF of the delay circuit 22 on the receiving side is controlled in the opposite phase to the ON / OFF of the delay circuit 9 on the transmitting side.

G3. Configuration of Delay Control Unit 4 FIG. 1 shows the configuration of the delay control unit 4 provided on the transmission side.

An initialization signal generated when the power is turned on is supplied to the input terminal 31, and a control pulse is supplied to the input terminal 32 once per field. Motion vector calculation circuit 3 at input terminal 34
Horizontal motion vector (5 bits of 2's
Complementary code) is supplied, and the set value (+14) is supplied to the input terminal 33. These motion vectors and set values are supplied to the selection circuit 36.

The motion vector from the input terminal 34 is supplied to one input terminal of the comparison circuit 37. The reference value (+15) from the input terminal 35 is supplied to the other input terminal of the comparison circuit 37. The comparison circuit 37 generates a comparison output which becomes high level when the motion vector has a magnitude equal to +15.

The comparison output of the comparison circuit 37 is supplied to the AND gate 38. The output signal of the AND gate 38 is supplied to the selection circuit 36. The selection circuit 36 selects the set value (+14) when the output of the AND gate 38 is at the high level, and selects the motion vector input when the output of the AND gate 38 is at the low level.

The 5-bit output data of the selection circuit 36 is supplied to the OR gate 39 and the delay circuit 40. The output of the delay circuit 40 is the OR gate 39
Is supplied to. The delay circuit 40 has a delay amount of 1 field. The OR gate 39 produces a low level (“0”) output only when horizontal movement is zero for two consecutive field periods.

The least significant bit (LSB) of the output of the selection circuit 36 is the inverter 4
1 and AND gate 44, and the output of inverter 41 is AN
It is supplied to one input terminal of the D gate 42. AND gate 42
The output of the latch 47 is supplied to the other input terminal of
The output of the D gate 42 is supplied to one input terminal of the AND gate 43.

The output of the OR gate 39 is supplied to the other input terminal of the AND gate 43. The output of the AND gate 43 is supplied to one input terminal of the OR gate 46.

The output of the latch 47 is supplied via the inverter 45 to the other input terminal of the AND gate 44 to which the least significant bit of the selection circuit 36 is supplied. Output of AND gate 43 and AND gate 44
Is supplied to the OR gate 46, and the output of the OR gate 46 is supplied to the latch 47.

A control pulse generated for each field from the input terminal 32 is supplied to the latch 47. The output of this latch 47 is AND
It is supplied to one input terminal of the gate 48. AND gate 48
A control pulse is supplied to the other input terminal of the AND gate 48, and the output of the AND gate 48 is supplied to the clock input terminal of the flip-flop 49.

An initialization signal generated when the power is turned on is supplied to the flip-flop 49 from the terminal 31 and is thereby cleared. The output signal on the positive side of the flip-flop 49 is the output terminal
At the same time as being taken out by 50, the output signal on the negative side of the flip-flop 49 is taken out by the output terminal 51.

The control signal taken out to one output terminal 50 is used to control the switch circuit 5 which defines ON / OFF of the delay circuit 9 (see FIG. 2) on the transmission side. That is, when the control signal output to the output terminal 50 is "1", the output terminal 8 of the switch circuit 5 is connected to the input terminal 7 and the delay is turned on. When the control signal is "0", the switch is switched. The output terminal 8 of the circuit 5 is connected to the input terminal 6 and the delay is turned off.

The control signal extracted to the other output terminal 51 is a signal for controlling the switch circuit 18 that regulates ON / OFF of the delay circuit 22 (see FIG. 3) on the receiving side, and is added to transmission data and transmitted. To be done. Switch circuit by this control signal 18
The control of is also similar to the control of the switch circuit 5 on the transmission side. Therefore, on the transmission side and the reception side, the delay ON / OFF is controlled in the opposite relationship.

An output signal on the negative side of the flip-flop 51 is returned to the input of the flip-flop 51, and the flip-flop 51 is inverted every time the clock is supplied. The output on the negative side of the flip-flop 51 is supplied to the other input terminal of the AND gate 38.

G4. Operation of the delay control unit 4 In the delay control unit 4 described above, the output signal of the AND gate 48 is "1".
Then, the state of the flip-flop 49 is inverted and the delay is inverted. The operation of the delay control unit 4 is represented by the flowchart shown in FIG.

First, the comparison circuit 37 checks whether the horizontal motion vector is equal to the reference value (+15) (step 61 in the flowchart of FIG. 4). In the case where the motion vector matches +15, when the delay is turned from OFF to ON, the motion becomes +16, which exceeds the motion range. In order to prevent this, when the motion vector is +15 and the previous delay is off, that is, when the output on the negative side of the flip-flop 49 is "1", the output of the AND gate 38 is set to "1", and the selection circuit 37 sets it. , Set value (+14) is selected, and the motion vector is forcibly changed from +15 to +14 (step 62).

Next, it is checked whether the delay was inverted the last time (meaning the previous field) (step 63). Since the previous state is held in the latch 47, the output of the latch 47 is "1" when the delay is inverted last time.

If there was a delay inversion last time, the LSB of the motion vector
It is checked whether (least significant bit) is "1" (step 64). When the LSB of the motion vector is "0", the output of the inverter 41 becomes "1", so the output of the AND gate 42 becomes "1".

Next, it is checked whether the motions of the previous field and the current field are both zero (step 65). When both movements are zero, that is, when the image is a still image, the output of the OR gate 39 becomes "0" and the output of the AND gate 43 also becomes "0". When it is not a still image, the output of OR gate 39 becomes "1", and AN
The output of the D gate 43 becomes "1".

Since the output of the AND gate 43 is supplied to the latch 47 via the OR gate 46, the output of the latch 47 becomes "1" in the current field. Therefore, the clock is supplied to the flip-flop 49 via the AND gate 48, the state of the flip-flop 49 is inverted, and the inversion of delay occurs (step 66).

The reason why the still image is inspected is to prevent malfunction due to noise or the like when a motion vector different from the actual motion of the image is obtained.

In other cases that cause delay inversion in the current field, the delay inversion does not occur last time, so when the output of the inverter 45 is "1" and the motion vector LSB is "1", that is, the AND gate.
It is when the output of 44 becomes "1". After step 63 to step 67, the process proceeds to step 66 of the delay inversion process.

G5. Specific Example of Operation of Delay Control Unit 4 The operation of the above-described embodiment of the present invention will be described more specifically with reference to FIGS. 5 and 6.

The example shown in FIG. 5 shows a case where the movements of the first and second fields are zero, a movement of 1 / (2fs) occurs in the third field, and the rest of the fourth and subsequent fields is stationary.

In the first field and the second field, there is no delay inversion last time, and the LSB of the motion vector is “0”.
Through steps 63 and 67 in the flow chart of the figure, the delay is not inverted in the current field.

In the third field, since there was no delay inversion last time and the LSB of the motion vector is "1", the process moves to step 66 through steps 63 and 67 in the flowchart of FIG. 4 and the delay in the third field Is inverted and the delay circuit 9 on the transmission side is inserted.

In the fourth field, since the delay is inverted in the previous third field, delay inversion is not performed even if the LSB of the motion vector is "1". That is, the step is the end step through steps 63 and 64 in FIG.

In the fifth field, since there is no delay inversion in the previous fourth field and the LSB of the motion vector is “0”, the delay inversion processing is not performed through steps 63 and 67.

In the example shown in FIG. 6, the movement of the first field and the second field is zero, the movement of 1 (2fs) occurs in the third field, and 1/2 for each field similarly in the fourth and subsequent fields.
The ratio of constant velocity panning in which (2fs) movement occurs is shown.

In the first field and the second field, there is no delay inversion last time, and the LSB of the motion vector is “0”.
Through steps 63 and 67 in the flow chart of the figure, the delay is not inverted in the current field.

In the third field, since there was no delay inversion last time and the LSB of the motion vector is "1", the process moves to step 66 through steps 63 and 67 in the flowchart of FIG. 4 and the delay in the third field Is inverted and the delay circuit 9 on the transmission side is inserted.

In the fourth field, there is a delay inversion in the previous third field, and since it is a motion of 1 / fs as compared with the second field, the LSB of the motion vector becomes “0”. Further, since the motions of the previous field and the current field are not zero, the delay inversion process is performed. That is, steps 63 and 6 in FIG.
After 4,65, the process proceeds to delay inversion step 66.

In the fifth field, there is delay inversion in the previous fourth field, and since the LSB of the motion vector is “0”, steps 63, 64, and 65 are executed in the same manner as described above, and step 66 of the delay inversion process is executed. Move to.

Delay inversion processing is also performed in the subsequent fields.

H. Effect of the Invention According to the present invention, in the case of a still image, by forcibly prohibiting the on / off inversion of the 1 / (2fs) delay circuit, the motion vector is erroneously obtained due to noise or the like. Even
It is possible to prevent malfunction of the reversal control.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of a main part of an embodiment of the present invention, FIG. 2 is a block diagram showing a configuration of a transmitting side of an embodiment of the present invention, and FIG. 3 is an embodiment of the present invention. FIG. 4 is a block diagram showing the configuration of the receiving side of FIG. 4, FIG. 4 is a flow chart for explaining the operation of the delay control section in one embodiment of the present invention, and FIGS. 5 and 6 are operation explanations of one embodiment of the present invention. FIG. 7 is a schematic diagram for explaining sub-sampling of the MUSE system to which the present invention can be applied, and FIG. 8 is a schematic diagram for explaining a range of motion detection of the MUSE system. FIG. 9 and FIG. 9 are schematic diagrams for explaining the delay control. Description of main symbols in the drawings 1: Digital high-definition television signal input terminal, 3: Motion vector calculation circuit, 4: Delay control unit, 5: Switch circuit,
9: Delay circuit.

Claims (2)

[Claims] 1. A band compression transmission apparatus for compressing and transmitting pixels of a digital television signal having a sampling frequency of 2f _s by sub-sampling, by performing block matching between consecutive frames of the digital television signal, Resolution is 1 /
A means for detecting a motion vector of (2fs), and 1 / (2f
s) delay means having a delay amount, selecting means for selectively extracting one of the output signal of the delay means and the digital television signal, subsampling means to which the output signal of the selecting means is supplied, A delay for generating a control signal for controlling the selecting means by determining the least significant bit "0" and "1" of the motion vector and detecting whether or not the state of the selecting means is changed in the previous field. The control means, means for judging whether or not the image is a still image by observing the motion vectors of two consecutive fields, and prohibiting the change of the state of the selecting means when the image is a still image; A band compression transmission device comprising a means for adding to the output signal of the sampling means. 2. A band compression transmission apparatus as set forth in claim 1, wherein the delay means is provided in common for the odd field and the even field of the digital television signal.