JPS62143584A - Transmission system using data compression - Google Patents

Abstract

PURPOSE:To transmit a data in a high compressibility by dividing a picture data into blocks, fitting the luminance value of a picture element by every block with the curved surface of a prescribed degree, and transmitting a parameter in which the square sum of a forecast error is set as the minimum. CONSTITUTION:An input digital television signal is supplied to a blocking circuit 2, and a picture F is divided forming a block B, and a picture (luminance) data is generated in a sequence of a numeric attached on the block B. The coordinate of the picture element having a luminance value Zi is expressed in xi and yi, and parameters a1-a10 in which the square sum of the forecast error against a true value Zi that is held by a forecast value Zi' calculated at a parameter identification part 3 becomes the minimum value are identified, and are transmitted to a reception side. At a restoration part 15 at the reception side, the luminance value is restored by the coordinate data of the block B and the parameters a1-a10, and the order of the data is recovered to that of the television signal with a scanning conversion circuit 16, thereby the data of a restored picture can be obtained.

Description

[Detailed Description of the Invention] [Field of Industrial Application] This invention compresses and transmits single image data such as television signals using high-efficiency codes, and forms a restored image on the receiving side. This invention relates to a transmission system using data compression.

[Summary of the invention]

The present invention is applied when transmitting image data, and divides input image data into blocks, and fits the brightness values of pixels in each block using a curved surface of a predetermined degree.
The parameter that minimizes the sum of squares of prediction errors during fitting is identified, this parameter is transmitted, and the receiving side restores the luminance value of the pixel in the block from the received parameter and block coordinate data. , data can be transmitted at a high compression ratio, a good restored image can be formed, and data compression and jV original processing can be easily performed.

C. Prior Art] A high-efficiency encoding method is known that compresses the data portion when transmitting image data with a large amount of information, such as a digital television signal. As one of the high-efficiency encoding methods, one television image is divided into two-dimensional regions (referred to as blocks) each consisting of a plurality of pixels.
Block encoding methods are known in which each block is encoded.

The conventional block encoding method calculates the average AV and standard deviation σ of a plurality of pixel data (luminance values) in a block, and allocates 1 bit to each pixel in the block.
A method is known in which the luminance value of (Av + σ) is encoded as "1" and the luminance value of (Av - σ) is encoded as "O", and the average value AV of each block and the encoded output of each pixel are transmitted. .

[Problem that the invention seeks to solve]

In the block encoding method described above, the restored image on the receiving side becomes a set of mosaics corresponding to the block size,
The pro7 distortion generated in the filling field between blocks becomes noticeable. Therefore, when it is desired to obtain a high-quality restored image, it is necessary to reduce the size of the frame r17.

By making the blocks smaller, the compression rate decreases, which has the disadvantage.

Therefore, it is an object of this invention to obtain high rF? The object of the present invention is to provide a transmission system that can obtain a good restored image at an iW rate.

Another object of the present invention is to provide a transmission system that has a simple configuration for identifying parameters and a simple configuration for performing restoration using received parameters.

[Means for solving problems]

This invention compresses the amount of input image data for each block consisting of a two-dimensional array of a plurality of pixel data,
In a transmission system that transmits compressed data and restores input image data on the receiving side, there is a blocking circuit 2 that converts the input image data into a data series having a sequence of 1J and 1J, and a luminance value of pixels in the block. is fitted by a curved surface of a predetermined order, and a parameter identification unit 3 identifies, for each block, parameters of the curved surface that minimizes the sum of squares of preselection errors during fitting.
, a restoring unit 15 that receives parameters and restores the luminance values of pixels in the block using block coordinate data, and a scan conversion circuit 16 that converts the pixel data into restored image data in the original order. It is something.

C Effect] Each block of pixels has an equal block size and is defined by common (X, y) coordinates. The luminance values of the pixels in this block are fitted by a predetermined curved surface of degree 3, and the parameters of each block that minimize the prediction error during fitting are transmitted. In the case of order 3, 10 parameters a1 to a
It is only necessary to transmit le, and the number of pixels in one block is M.
If this is done, a compression ratio of (10/M>) can be achieved.In addition, since the fitting is performed using a curved surface, the restored image quality can be improved compared to encoding with two strands.Furthermore, The coordinate data constructed based on the coordinates of one block is common to each block, and it is possible to easily identify parameters and form a restored image.

〔Example〕

An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 shows the configuration of a transmission system consisting of a transmitting side and a receiving side. In FIG. 1, ``■'' is an input terminal for a digital television signal.

An input digital television signal is supplied to a blocking circuit 2 which converts the data order from television scan order to block order. In Figure 2, F is
One television image is shown. In the television image F, pixel data is generated from the left end to the right end of the line, and in the vertical direction, pixel data for each line is generated from the top to the bottom. As shown in FIG. 2, the blocking circuit 2 divides the image F in the 4if direction into five, for example.
One block B is formed by dividing, for example, into five in the horizontal direction. Image (brightness Iff) data is generated from the blocking circuit 2 in the order of the numbers assigned to each block B. Overlapping the boundaries of adjacent blocks B is effective in preventing block distortion.

Figure 3 shows one block B, which includes (4N x 1) pixels in the horizontal direction and (2N x 1) pixels in the vertical direction, for a total of M pixels CM-(4N x 1). x (2N-t-1
)] pixels are included. The position of each pixel within this block B is expressed by coordinates on the X and y axes with the origin O at the center. Let the brightness value of the pixel located at the upper corner of the left end in block B be ZI, and from this data, the brightness value of the pixel data located in the horizontal direction is 22. Z3..., and
Let ZM be the luminance value of pixel data located at the lower right corner of block B. The coordinates of a pixel having a luminance value of Z are expressed as (x+, Yl).

The output signal of the blocking circuit 2 is supplied to the variable identification section 3. The distribution of all luminance values included in one block, θ, can be approximately expressed by the (x, y) coordinates shown in FIG. In other words, when the data in a block is fitted to a curved surface having three poles, the predicted brightness value 2 is expressed by the following equation.

a+X to Z;"az X+'Yl"a3x; VI2
184 y% 10a5X(,'+ab Xi yi+a
v3'+"-as χ, → a9 Vi +aIa In the ten equations, (X;, y,) indicates the (17 position in one block) of the pixel with the brightness value Z, and a1 to alQ are parameters. be.

In the parameter identification section 3, the predicted value 2. Parameters a1 to a1o are identified that minimize the sum of squares of prediction errors that correspond to the true value Z. Parameters a, ~a of each block from the parameter identification unit 3
lo is taken out to the output terminal 9 and sent to the receiving side.

The number of appended pins of luminance (ilY) is 8 hits, and the number of bits of parameters a1 to alQ is 8 bits in one row.

The input terminal 11 on the receiving side receives the parameter a from the transmitting side.
1~a. is received and supplied to the restoration unit 15. In the restoration unit 15, the coordinate data of block B and parameters a,
~a16, the luminance value is restored. Restored data from the restorer 15 is supplied to a scan conversion circuit 16. The scan conversion circuit 16 returns the data order to that of the television signal, and the restored image data is obtained at the output terminal 17.

In this example, M(=(4N
)X (2N11) ] pieces of data are compressed into 10 parameters a1 to alo, allowing for significant compression. Parameter a determined by parameter identification unit 3
Identification of 1 to alo will be explained below.

When predicting the brightness values 21 to ZM of M pixels in block B by fitting a curved surface of degree 3, and assuming that the prediction error is e, ez...e, 4, the brightness values 21 to ZM of M pixels in block B are calculated using the following matrix It is shown by calculation.

;　Zl　　： ・　　ZM　　’ L--” The above matrix operation can be rewritten as follows.

'1E=W-a+10 In other words, l is an M-order hector, W (M rows and 10 columns) matrix, al is a 10-order vector, and e is an M-order vector. The parameter a1 that minimizes the sum of squares of errors determined by the least squares method is expressed by the following equation.

a+ -(W''W)-'W''-Coordinates of each luminance value in the Z block (Xi, X2...X)1.
Vl, y2. . . . y, 4) becomes fixed data once the value of N corresponding to the pronoccysis is determined. Therefore, (
The term WTW)-'W'' (coordinate data) is i1M fixed data for all blocks, and can be generated by, for example, a ROM.

The matrix of (WT W )-1 is [(10 rows and 4 columns)×(
Since it is the product of [(10 rows and 10 columns)], it becomes (10 rows and 10 columns), and the matrix of (WTW)-'W'' is the product of Because there is (row 10, column M)
becomes a matrix of Therefore, the formula for determining the parameter at using the least squares method is as follows.

Regarding fitting of data within a block, for ease of understanding, an example will be described in which a one-dimensional block is fitted with a straight line. 7 on the in if 1 block is the same
As shown in FIG. 4, an X coordinate is formed with the center pixel as the origin. Brightness value of each pixel 2. ．．
22. . . . When Z7 has the change shown in FIG. 4, the luminance value within the block is fitted by a straight line (ax - b) shown by a broken line. Therefore, the brightness value of each pixel is
It is expressed by the following formula.

- the parameters a and b are identified by the least squares method. It is not a straight line, but a quadratic curve (ax” fb
When fitting x0C), 3 of a, b, c
It is necessary to identify these parameters. In this invention, since it is a two-dimensional block, a curved surface is used instead of a straight line or a curved line.

FIG. 5 shows an example of the parameter identification section 3.

The luminance value Z from the blocking circuit 2 is input to the input terminal indicated by 20.
, is supplied. This brightness value Z1 is divided into 10 multiplier circuits 2
1.22. ... is supplied to each of the 30. Multiplication circuit 21.22. . . 30 contains coordinate data Wk(1), W of luminance 4FiZk from a ROM (not shown).
k(2), . . . W11Qω are supplied.

Each multiplication circuit 21, 22 . . . 30, each output data is sent to the adder circuits 31, 32 . . . . is supplied to one input terminal of 40. Addition circuit 31. ．． 32°...4
Each output data of 0 is stored in registers 51, 52, . . . 6
0 and output terminals 71, 72 .・・・
・Taken out at 80. A pair of adder circuit and register constitutes an integrator circuit. Registers 51, 52.
. . . 60 is reset every time the calculation of M brightness values of block (1) is completed.

When M luminance values are supplied to the input terminal 20, the output data obtained at the output terminal 71 is [Wl(1)Zl +W
z(1)L 10−−−+WM (1)ZM =a.

], and the parameter a1 is generated. Similarly, output terminals 72, 73 . . . 80, each has a barometer a2. a3. ... Ago is taken out.

The restoration unit 15 provided on the receiving side has the configuration shown in FIG. 6. Although not shown in the figure, the restoring unit 15 is provided with a ROM, and different coordinate data are stored in the ROM.
Occurs twice. While this coordinate data occurs M times,
Each of the input terminals 91, 92°...100 receives one parameter a I=a2+...al
e is supplied.

Parameter al+ aZ+...al. is the multiplication circuit 1
01°102, . . . 109. For example, coordinate data related to the coordinates (xl, y,) of the i-th pixel
[x I3. x 12 yl, xl y I
2. ...yl] force (multiplying circuit 101..10
2. ...109 respectively. Multiplication circuit 10
1,102. . . . 109 are sent to adder circuits 112, 113 . ...120 is added. Therefore, the output terminal 121 derived from the adder circuit 120 at the final stage receives the restored luminance value 2.
is obtained.

Zi = a + Xt ”-az X%V1shia2 X
i )'+2 24Y+ -as X+'+ab Xr
V, 10a, )', 2-1-a @X, +
aq yi +3g.

An example (experimental result) of the standard deviation of the restoration error when the size of one block is changed is shown below.

Note that the present invention can be similarly applied to the case where an n-dimensional curved surface having an order other than 3 is used.

〔Effect of the invention〕

This invention performs fitting of a curved surface of a previously prepared order for each block, and transmits only the curved surface of the above-mentioned order that provides the best fitting, that is, the parameters that minimize the sum of squared errors. Therefore, according to this invention, a very large compression ratio can be obtained, and the
Brightness values are expressed by various curved surfaces, and image quality can be significantly improved compared to conventional binary block encoding.

Furthermore, in this invention, when identifying parameters, common coordinate data is used between blocks, so for example, RO
By generating the coordinate data using M, it is possible to easily configure the parameter identification section using only multiplication and addition processing.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing the configuration of an embodiment of the present invention.
Figures 2 and 3 are route maps for explaining block formation and one block, Figure 4 is a route map for facilitating understanding of fitting according to the present invention, and Figure 5 is an embodiment of the present invention. FIG. 6 is a block diagram of an example of the parameter identification section in FIG. 8, and FIG. 6 is a block diagram of an example of the restoration section in one embodiment of the present invention. Explanation of main symbols in the drawings 2: Niblocking circuit, 3: Parameter identification section, 15
: Restoration unit, 16: Scan conversion circuit. Agent Patent Attorney Tadashi Sugiura Chibon Systemhe 17011. Figure 3 (xl) Naoe Abu Motel
(x6) Figure 4

Claims (1)

[Claims] The amount of input image data is compressed for each block consisting of a two-dimensional array of a plurality of pixel data, the compressed data is transmitted, and the input image data is restored on the receiving side. A transmission system comprising means for converting the input image data into a data series having the order of the blocks, and fitting the luminance values of pixels in the blocks by a curved surface of a predetermined order, and squares the prediction error at the time of the fitting. means for identifying, for each block, a parameter of the curved surface that minimizes the sum; means for receiving the parameter and restoring the luminance value of a pixel in the block using coordinate data of the block; 1. A transmission system using data compression, comprising means for converting the data into restored image data in the original order.