JP2778629B2 - High efficiency coding device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial Application Field] The present invention relates to an image
High-efficiency coding that compresses the number of bits per pixel of data
Related to the device. [Related Art] As a method of encoding a television signal, a transmission band is narrowed.
Average number of bits per pixel or
Several ways to reduce the sampling frequency are known.
Have been. As encoding methods for lowering the sampling frequency,
Subsampling reduces the image data to 1/2
Sampling point and subsample used for interpolation
Indicates the position of the interpolation point (that is, whether
Indicates whether to use the data of the sub-sampling points)
One that transmits the data has been proposed. Coding method for reducing average number of bits per pixel
As one of them, DPCM (differential PCM) is known
You. DPCM has a high correlation between pixels of the television signal.
Focusing on the small difference between adjacent pixels,
The difference signal is quantized and transmitted. Coding method for reducing average number of bits per pixel
Another thing is that a screen of one field is
Subdivided into blocks and the representative point pixels and blocks
Some transmit the deviation of the level distribution of the data in the
You. [Problems to be solved by the invention] Reduce sampling frequency by using subsampling
The encoding method to be used is that the sampling frequency is 1/2
, There is a possibility that aliasing may occur. DPCM has the problem that errors propagate to subsequent decoding.
Was. The method of encoding in block units is based on the
There is a disadvantage that block distortion occurs at the boundary. Therefore, the generation of aliasing distortion of the above-described conventional technology is considered.
No problems such as generation of errors, propagation of errors, occurrence of block distortion, etc.
A high efficiency coding apparatus has been proposed. The present applicant is described in JP-A-59-266407.
Of a plurality of pixels contained in a two-dimensional block
Dynamic range defined by maximum and minimum values
And perform encoding adapted to this dynamic range.
A highly efficient coding device has been proposed. In addition, Japanese Patent Application No. 60-
Multiple fields as described in 232789
The three-dimensional block formed from the pixels included in
A coding method adapted to the dynamic range has been proposed
I have. With coding methods adapted to these dynamic ranges
Is a block of a still image and a block of a moving image.
Have the same quantization step. Further, it is described in the specification of Japanese Patent Application No. 60-268817.
As you can see, the maximum distortion that occurs when performing quantization is constant.
The word length (number of bits) changes according to the dynamic range.
Variable-length coding methods have been proposed. In this variable-length coding method, still image
The maximum distortion is the same between the block and the moving block
It has been. However, gradually increase the quantization step to increase the compression ratio.
If it is set higher, the deterioration of block distortion etc.
Occurs. The reason for this is that images with relatively fast movement
In this case, the persistence characteristics of the CRT and the integration effect of the human eye
In the case of a still image,
This is because even the details of the image can be recognized. The present invention considers this visual characteristic and
The present invention is to improve a high-efficiency coding apparatus using a clock. Immediately
In the present invention, the amount of motion of an image is detected in block units.
Then, a motion code indicating the amount of motion is generated, and the motion code
Encoding using a quantization step adapted to the amount of motion
The high efficiency mark that can further increase the compression ratio
A decoding device. [Means for Solving the Problems] The present invention is directed to a plurality of continuous digital image signal filters.
Form a block consisting of regions belonging to each of the fields
Means and the amount of motion of the image is determined for each block, and the motion code is determined.
The amount of motion that is generated and the quantization step for quantizing the pixel signal of the block are activated.
Corresponding to the amount of motion given by the
If it is large, the quantization step width will also be large,
Quantization means determined adaptively, quantized data and motion code output from the quantization means
High efficiency coding characterized by comprising means for transmitting data
Device. [Operation] The television signal is divided into horizontal and vertical directions and time.
Because it has a three-dimensional correlation in the direction,
Now, the level of the pixel data contained in the same block
The change width is small. Therefore, the pixel data in the block is shared.
The dynamics of the data after removing the minimum level
The number of quantization bits less than the original number of quantization bits
, The quantization distortion hardly occurs. quantum
Data transmission bandwidth by reducing the number of
The width can be narrower than the original. Also the human eye
Considering the perceptual characteristics, the amount of coding
The nesting step has been changed to increase the compression ratio,
Deterioration of the image of the stationary part can be prevented. Hereinafter, embodiments of the present invention will be described with reference to the drawings.
I do. This description is made in the order given below.
You. a. Configuration on the transmission side b. Configuration on the reception side c. Block and blocking circuit d. Motion amount detection circuit e. Dynamic range detection circuit f. Quantization circuit g. Modification a. Configuration on the transmission side The entire structure of the transmitting side (recording side) of the present invention
It is shown as. For example, one input terminal
Digital television with sample quantized to 8 bits
Input signal. This digital television signal
Is supplied to the blocking circuit 2. The input digital television
Signal is continuous for each block, which is the unit of coding
Is converted to The output signal of the blocking circuit 2 is
Output circuit 3 and the dynamic range detection circuit 4
You. The motion detection circuit 3 is a three-dimensional block (in this example,
Indicates the amount of motion within 6 lines x 6 pixels x 3 frames)
For example, a circuit that generates a 2-bit motion code SJ
You. As the amount of movement increases, the movement code SJ becomes [00]
It changes to [01] [10] [11]. Dynamic range detection
The output circuit 4 has a maximum value MAX, a minimum value MIN, a dyna
Mic range DR is detected. Block from block circuit 2
The input data converted to the lock order is supplied to the subtraction circuit 5.
And the minimum value MIN is removed in the subtraction circuit 5.
Data PDI is formed. The motion code SJ from the motion amount detection circuit 3 is
Is supplied to the loop determination circuit 6. Quantization step determination circuit 6
Is a motion code SJ with a nonlinear relationship matching the visual characteristics.
Is determined according to. This quantization step
Is supplied to the bit number determination circuit 7. Bit count
The constant circuit 7 includes a signal from the dynamic range detection circuit 4.
The dynamic range DR is supplied. Bit number determination circuit 7
The dynamic range DR is divided by the quantization step Δ
The result of the division is converted to a binary number,
Is done. The number of bits Nb is supplied to the quantization circuit 8. The data PDI from the subtraction circuit 5 is supplied to the quantization circuit 8.
It is. In the quantization circuit 8, the dynamic level of each block is
The encoding is performed with a variable number of bits adapted to the dynamic range DR.
That is, the quantization circuit 8 receives the video from the bit number determination circuit 7.
Nb are supplied, and the data PDI after minimum value removal is bit
Quantized by the number Nb. The command from the quantization circuit 8
The mode signal DT is supplied to the framing circuit 9. In this embodiment, the motion code SJ, the dynamic
To transmit DR, minimum MIN, and code signal DT.
I have. These data are supplied to the framing circuit 9,
Converted to transmission data. As the form of the transmission data,
Motion mode SJ, minimum MIN, dynamic range DR and
Independent error for each data part consisting of code signal DT
The correction code is encoded and the parity of each error correction code is
The one that transmits the data with the address added can be used. Also the code
Motion code SJ other than signal, dynamic range DR and
Independent error correction codes are applied to each of the small values MIN
May be. Furthermore, motion code SJ, dynamic range DR
Of error correction code common to both MIN and MIN
And the parity may be added. Framing
Transmission data is taken out at an output terminal 10 of the circuit 9. Illustrated
The transmission data from the framing circuit 9 is
Transmitted (or recorded on a recording medium). b. Configuration on the receiving side FIG. 2 shows the configuration on the receiving (or reproducing) side. Input end
The received data from the slave 21 is supplied to the frame decomposition circuit 22.
You. The code signal DT and the additional code
Code MIN, DR and motion code SJ are separated,
-Correction processing is performed. The code signal DT from the frame decomposition circuit 22 is used as a decoding circuit.
Supplied to 25. Motion code SJ determines quantization step
Supplied to the path 23 and supplied from the quantization step determination circuit 23.
The conversion step Δ is supplied to the bit number determination circuit 24,
The mix range DR is supplied to the bit number determination circuit 24. This
These quantization step determining circuit 23 and bit number determining circuit
24 is the quantization step determination circuit 6 on the transmission side and the number of bits
The configuration is the same as that of the decision circuit 7. The decoding circuit 25 stores the number of bits Nb together with the code signal DT.
The supplied and variable-length coded code signal DT is decoded.
You. The decoding circuit 25 is the inverse of the quantization circuit 8 on the transmitting side.
Performs processing and outputs a code signal and corresponding level
A signal is generated. The output signal of the decoding circuit 25 is supplied to the adding circuit 26.
The data after the removal of the minimum level of 8 bits in the decoding circuit 25
Data DTI is restored as the representative level, and this data and 8
The minimum value MIN of the bit is added by the adder circuit 26 to obtain the original image.
The raw data is decrypted. The output data of the adder 26
Is supplied to the clock decomposition circuit 27. The block decomposition circuit 27
Contrary to the blocking circuit 2 on the transmitting side, the order of the blocks is restored.
Signal data in the same order as the scanning of the television signal
It is a circuit for performing. Output terminal of block decomposition circuit 27
The original television signal decoded at 28 is obtained. c. Block and blocking circuit Referring to FIG.
Will be described. In FIG. 3, reference numeral 14 denotes each of three frames.
One frame consisting of the two-dimensional areas 14A, 14B, 14C belonging to the frame
The solid line indicates the lock in the odd field.
, And the dashed line indicates the line of the even field. Each
The six pixels in each of the six lines of the frame
Thus, the areas 14A, 14B, and 14C of (6 lines × 6 pixels) are configured.
Is done. Therefore, one block is (6 × 6 × 3 = 108)
Pixels. Original digit included in one block
The total number of bits of the television signal is (108 × 8
Bits = 864 bits). The number of quantization bits of the code signal is
The less the better. However, it did not increase the quantization distortion.
For this reason, the number of quantization bits must not be reduced too much.
When the number of quantization bits is 8 bits,
There are 256 levels (0 to 255). But things
In the stationary part except the non-stationary part such as the body contour, one block
Pixel level distribution is concentrated in a fairly narrow range
doing. In the case of a television signal, a three-dimensional one-block
Since each pixel in the block has a correlation,
The dynamic range DR does not become very large,
As a high value, it is enough to consider the 128th place. FIG. 4 shows an example of the configuration of the above-described blocking circuit 2.
You. Frame memory 15A, 15B, 15C cascade connection to input terminal 1
Has been continued. The pixel data of the current frame Fn + 2 runs
Is supplied to the frame conversion circuit 16A, and is sent from the frame memory 15A.
The pixel data of the frame Fn + 1 is supplied to the scan conversion circuit 16B.
Earlier frame Fn from frame memory 15A
Are supplied to the scan conversion circuit 16C. The scan conversion circuit 16A includes one frame, as shown in FIG.
Converts the order of data in a frame to the order of each block area
I do. Similarly, the other scan conversion circuits 16B and 16C also
The order of the data in the block to the order of each block area
You. The output data of the scan conversion circuit 16A passes through the delay circuit 17A.
The output signal of the scan conversion circuit 16B
Is supplied to the synthesizing circuit 18 via the delay circuit 17B. delay
The circuit 17A includes 72 pixel data included in the two regions and
The delay circuits 17B have the same delay amount, and the delay circuit 17B is included in one region.
It has a delay amount equal to rare 36 pixel data. Also,
The synthesis circuit 18 includes a delay circuit and a switch circuit.
ing. As shown in FIG. 5B, the output terminal 19 of the synthesis circuit 18
Included in each of three consecutive frames Fn, Fn + 1, Fn + 2
The pixel data of the areas 14A, 14B, 14C
You. That is, the output terminal 19 is converted to the block order.
Output data is obtained. d. Motion amount detection circuit FIG. 6 shows an example of the motion amount detection circuit 3. In FIG.
Here, each of 31A, 31B, and 31C has three frames in one block.
Input terminals to which image data of frames Fn, Fn + 1 and Fn + 2 are supplied
I am a child. This input data is supplied to the above-mentioned blocking circuit 2
Output data is parallelized block by block
It is formed. From the image data of each frame and the terminal 31D
Threshold data Tf and the reset pulse P from terminal 31E
R is supplied to the motion amount detection circuits 32A, 32B and 32C, respectively.
You. The motion amount detection circuit 32A detects the area 14A of the frame Fn and the frame Fn.
The motion of the image is detected between the areas 14B of the frame Fn + 1. Motion
The amount detection circuit 32B includes the region 14A of the frame Fn and the frame Fn.
The motion of the image is detected between the +2 areas 14C. Motion detection
The circuit 32C includes an area 14B of the frame Fn + 1 and a frame Fn +
The motion of the image is detected between the two areas 14C. These moves
The amount detection circuits 32A, 32B and 32C filter out the input image data.
In FIG. 6, the motion amount detection circuit
A specific configuration is shown for the road 32A. Motion detection
Output signals of multiple bits of output circuits 32A, 32B, 32C. Addition times
The output signal of the addition circuit 33 is supplied to the judgment circuit 34.
Supplied. The 2-bit motion code SJ is obtained from the decision circuit 34.
It is taken out to the output terminal 35. The motion amount detection circuit 32A includes a subtraction circuit 36, an absolute value conversion circuit 37,
It comprises a comparison circuit 38 and an aggregation circuit 39. Subtraction circuit
36 and the absolute value conversion circuit 37, between the area 14A and the area 14
Of the pixel level difference (frame difference) of the corresponding device of
The value is formed. The total value of the frame difference is calculated by the comparison circuit 38.
Is compared with the threshold data Tf. The level relation between the absolute value of the frame difference and the threshold data Tf
The binary comparison output corresponding to the person in charge is supplied to the summation circuit 39.
You. The counting circuit 39 counts the number of pixels exceeding the threshold value Tf.
Measure. The counting circuit 39 includes a reset pulse for each block.
PR is supplied. As described above, the motion amount between the regions 14A and 14C is the motion amount.
The movement between the regions 14B and 14C is detected by the detection circuit 32B.
The movement amount is detected by the movement amount detection circuit 32C. these
The motion detected by each of the motion amount detection circuits 32A, 32B, 32C
The amount is supplied to the adding circuit 33. Therefore, the addition circuit 33
The output signal has a value corresponding to the motion amount of the entire block.
You. The output signal of the adder circuit 33 is supplied to the decision circuit 34.
And the determination circuit 34 determines the amount of motion of the block.
And a 2-bit motion code SJ is generated. This movement
Block SJ is used to convert a block image into a still image
Image, medium-motion image, or large-motion image
Or shown. As a method of detecting the amount of motion, in addition to this embodiment,
The absolute value of the same pixel difference between frames
The amount of motion or the absolute value of the same pixel in framing
Use a method that uses the integrated value of the frame difference as the amount of motion.
Can be. e. Dynamic range detection circuit FIG. 7 shows an example of the configuration of the dynamic range detection circuit 4.
This is shown. The input terminal indicated by 41 has a blocking circuit
2. As described above, areas that require encoding for each block
The image data of the area is sequentially supplied. From this input terminal 41
Are supplied to the selection circuit 42 and the selection circuit 43.
You. One selection circuit 42 is connected to the input digital television.
Between the pixel data of the signal and the output data of the latch 44.
Select the higher level and output. The other choice times
The path 43 is the pixel data of the input digital television signal.
Between the output data of the latch 45 and
And output. When the output data of the selection circuit 42 is supplied to the subtraction circuit 46
Both are taken into the latch 44. Output data of the selection circuit 43
Data is supplied to the subtraction circuit 46 and the latch 48, and
It is taken into the h 45. Latches 44 and 45
Is supplied from the control unit 49. The control unit 49 has an input
Sampling clock synchronized with the digital television signal.
Timing signals such as clock and synchronization signals are supplied from terminal 50.
You. The control unit 49 latches on the latches 44, 45 and 47, 48.
The pulse is supplied at a predetermined timing. At the beginning of each block, the contents of latches 44 and 45 are initialized.
Is determined. Latch 44 is initially set to “0” data
The data of all "1" is initialized in the latch 45.
You. In the pixel data of the same block sequentially supplied
The maximum level is stored in the latch 44. In addition,
Of the pixel data of the same block
Is stored in the latch 45. Detection of maximum level and minimum level per block
Is completed, the output of the selection circuit 42
Levels arise. On the other hand, the output of the selection circuit 43
A minimum level of noise occurs. Detection for one block
When finished, the latches 44 and 45 are initialized again. The output of the subtraction circuit 46 has the maximum level from the selection circuit 42.
MAX and the minimum level MIN from the selection circuit 43 are subtracted.
The dynamic range DR of each block is obtained. these
Dynamic range DR and minimum level MIN
A latch pulse from the latch 49 causes the latches 47 and 48 to
Latched. The output terminal 51 of the latch 47 is
The dynamic range DR is obtained, and the output terminal 52 of the latch 48 is
Then, the minimum value MIN of each block is obtained. f. Quantization circuit The quantization circuit 8 can be adapted to the dynamic range DR.
The variable length encoding is performed respectively. FIG. 8 shows an example of the quantization circuit 8.
Here is an example. In FIG. 8, the ROM indicated by 55 has the minimum value
The compressed pixel data PDI (8 bits)
Data conversion table for converting to
I have. The number of bits Nb from the input terminal 56 is
Pixel data PDI from input terminal 57 is used as an address signal.
Supplied. In ROM55, the data conversion table is determined by the number of bits Nb.
Is selected and the code signal DT of Nb bits is
Be sent out. This number of bits Nb is quantized as described above.
With the quantization step Δ set by the step determination circuit 6,
Divide the dynamic range DR, convert it to binary, and find
It was a thing. Therefore, the number of bits Nb is determined by the quantization step.
And the dynamic range DR. In the still image block diagram, the finest quantization
For example, the maximum distortion E is 4 (quantization step Δ = 2E = 8)
It is said. In this case, the number of bits Nb is 0 to 5 bits.
Range. Therefore, the command output from ROM 55
The effective bit length changes in the code. Frame circuit
At 9 a valid bit is selected. FIG. 9 shows the case where (Δ = 8)
Used to describe variable-length coding
You. This encoding replaces the pixel data with the minimum value removed.
This is the process of converting to the table level. That occur during this quantization.
The maximum allowable quantization distortion (referred to as maximum distortion) is
As described above, it is set to 4. FIG. 9A shows that the dynamic range DR is (maximum value MAX
And the minimum value MIN) is 8. (DR = 8)
In this case, the central level 4 is set to the representative level L0,
(Distortion E = 4). That is, when (0 ≦ DR ≦ 8),
The middle level of the dynamic range is the representative level.
Therefore, there is no need to transmit the quantized data. Follow
Thus, the required 8-bit number Nb is 0. On the receiving side
Is the minimum value of the block MIN and the dynamic range DR?
The decoding is performed using the representative level L0 as a restoration value. FIG. 9B shows the case of (DR = 7), where the representative level is
(L0 = 4) (L1 = 13), and the maximum distortion E is 4
Become. Since there are two representative levels L0 and L1, (Nb = 1)
Becomes In the case of (9 ≦ DR ≦ 17), (Nb = 1)
You. The maximum distortion E is smaller as the dynamic range DR is smaller.
Become. FIG. 9C shows the case of (DR = 35), where the representative level is
(L0 = 4) (L1 = 13) (L2 = 22) (L3 = 31)
(E = 4). There are four representative levels L0-L3
Therefore, (Nb = 2). (18 ≦ DR ≦ 35)
Is (Nb = 2). In the case of (36 ≦ DR ≦ 71), eight representative levels (L0 to
L7) is used. FIG. 9D shows the case of (DR = 71)
And the representative level is (L0 = 4) (L1 = 13) (L2 = 22) (L3
= 31) (L4 = 40) (L5 = 49) (L6 = 58) (L7 = 67)
Each is determined. To distinguish eight representative levels L0-L7
, (Nb = 3). In the case of (72 ≦ DR ≦ 143), 16 representative levels (L0
To L15) are used. FIG. 9E shows the case of (DR = 143)
And the representative level is (L8 = 76) (L9 = 85) (L10 = 9
4) (L11 = 103) (L12 = 112) (L13 = 121) (L14 = 13)
0) (L15 = 139) (L0 to L7 are the same as the above values)
Can be To distinguish between 16 representative levels (L0 to L15)
, (Nb = 4). In the case of (144 ≦ DR ≦ 287), 32 representative levels (L0
To L31) are used. Fig. 9F shows the case of (DR = 287)
And the representative level is (L16 = 148) (L17 = 157) (L18
= 166) (L19 = 175) ... (L27 = 247) (L28 = 25)
6) (L29 = 265) (L30 = 274) (L31 = 283) (L0-L15
Is the same as the above value). 32 representative levels
(Nb = 5) to distinguish (L0 to L31). Real
In this case, the input pixel data is quantized by 8 bits.
Therefore, the maximum value of the dynamic range DR is 255,
It is not quantized to the table level (L28 to L31). For blocks with image motion, the amount of motion and
The quantization step Δ is made larger corresponding to the shape. One
In other words, coarse quantization is performed,
Thus, the maximum value of the number of bits Nb is made smaller. Therefore,
Regardless of the amount of motion, a constant quantization step (for example,
For example, the compression rate can be increased as compared with the method using 8). TV signal in one block is horizontal and vertical
Two-dimensional direction and three-dimensional correlation in time direction
In the constant part, it is included in the same block
The change width of the data level of the pixel to be changed is small. Follow
The minimum level MIN shared by the pixel data in the block
After removing the data PDI dynamic range
Quantization with a smaller number of quantization bits than the number of quantization bits
However, almost no quantum delay distortion occurs. Number of quantization bits
The transmission bandwidth of the data by reducing
Can be narrower. g. Modifications The present invention is not limited to variable-length coding
It can also be applied to coding schemes. Fixed-length coding
Then, the dynamic range DR of the block is
Divided into a number of level ranges determined by the number, and after removing the minimum value
The code signal corresponding to the level range to which the
Is done. Therefore, the quantum for coding a still block is
When the number of coded bits is set to 4 bits,
The number of bits is made smaller. For example, for a stationary block, as shown in FIG.
And the dynamic range DR is divided into 16 levels
Determines the level range to which the data PDI after minimum value removal belongs
And generates a code signal indicating the determined level range.
Live. On the other hand, for blocks with a small amount of motion
Is the same dynamic as in FIG. 10A, as shown in FIG. 10B.
Dynamic range DR, the dynamic range DR
It is divided into eight level ranges and contains the data PDI.
The bell range is determined. Therefore, for stationary blocks
Assuming that the quantization step of Δ is Δ, in the example of FIG.
The quantization step for blocks with few
Done. In the above-described quantization, as is apparent from FIG.
Namic range is equalized by quantization step Δ or 1 / 2Δ
Divide and represent the representative levels L0, L1, ...
.. Are used as values at the time of decoding. This encoding method
The method can reduce quantization distortion. On the other hand, each level of the minimum level MIN and the maximum level MAX
Pixel data that has
You. Therefore, in order to increase the number of encoded codes having an error of 0,
As shown in FIG. 11, the dynamic range DR is set to (2 ^m −
1) (where m is the number of quantization bits)
Bell MIN as representative minimum level L0 and maximum level MAX as representative
The maximum level L3 may be used. The example in Figure 11 is for simplicity.
2 shows a case where the number of quantization bits is 2 bits. In the above description, the code signal DT and the dynamic range
The DR, the minimum value MIN, and the motion code SJ are transmitted. Only
Quantum instead of additional code and dynamic range DR
Or the maximum distortion may be transmitted. Also, one frame of data frame memory, line delay
By combining a delay circuit and a sample delay circuit,
You may take out simultaneously. [Effect of the Invention] According to ADRC coding, the amount of data to be transmitted is
Data can be sufficiently reduced compared to
Can be. In addition, ADRC encoding has a variation range of the luminance level.
In the small stationary part, the original pixel data is roughly
Can be completely restored, and there is almost no deterioration in image quality.
There is a point. Furthermore, ADRC coding has a dynamic range
Since it is determined corresponding to each block, edges with large variation
The response in a transient part such as a jig is improved. According to the present invention, the amount of movement is large in consideration of human visual characteristics.
Static or dynamic blocks compared to the block quantization step
To reduce the quantization step for smaller blocks
Code transmission without deteriorating the quality of the restored image on the receiving side.
The number of bits of the signal can be reduced. Therefore, a higher compression ratio
Can be done.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a block diagram of an embodiment of the present invention, FIG. 2 is a block diagram showing a configuration on a receiving side, and FIG. 3 is a block diagram showing a unit of a coding process. FIG. 4 and FIG. 5 are schematic diagrams for explaining an example of the configuration of the blocking circuit and a description of the blocking circuit, FIG. 6 is a block diagram of an example of the motion amount detection circuit, and FIG. Is a block diagram of an example of a dynamic range detection circuit, FIG. 8 is a block diagram of an example of a quantization circuit, FIG. 9 is a schematic diagram used to explain an example of quantization,
FIG. 10 and FIG. 11 are schematic diagrams used to explain another example of quantization and still another example, respectively. Description of main reference numerals in the drawings 1: input terminal of digital television signal, 2: blocking circuit, 3: motion amount detection circuit, 4: dynamic range detection circuit, 6: quantization step determination circuit, 7: bit number determination Circuit, 8: quantization circuit, 9: framing circuit.

Claims (1)

(57) [Claims] Means for forming a block consisting of regions belonging to each of a plurality of continuous fields of the digital image signal; motion amount detecting means for determining a motion amount of an image for each of the blocks to generate a motion code; Quantizing means adaptively determined such that a quantization step for quantizing a pixel signal corresponds to a motion amount given by the motion code, and increases when the motion amount is large; And a means for transmitting the motion code.