Message compression and recovery process

FIELD: electrical communications; data processing including reduction of data redundancy.

SUBSTANCE: proposed process includes similar way of generation of random quadrature matrix measuring m x m items and k random key matrices measuring N x m and m x N items on sending and receiving ends. Then k matrices of quantum readings of motionless gray-level video picture measuring M x M items are formed from k motionless gray-level video pictures which are then converted into product of three following matrices: random rectangular matrix measuring N x m items, random square matrix measuring m x m items, and random rectangular matrix measuring m x N items; in the process items of rectangular matrix measuring N x m items are transferred to communication channel. On receiving end k matrices of recovered quantum readings of motionless gray-level video pictures measuring M x M items are formed around random matrix measuring N x m items received from communication channel, as well as around random quadrature matrix measuring m x m items, and random rectangular matrix measuring m x N items, and motionless gray-level video pictures are produced from mentioned k matrices of recovered quantum readings.

EFFECT: enhanced data transfer speed at desired quality of recovered messages.

4 cl, 24 dwg

 

The invention relates to the field of telecommunications, namely, to methods of digital computing and data processing with the reduction of the redundancy of the transmitted information. The proposed method can be used for the transmission of a fixed video via digital communication channels and belongs to the class of encodings-based recovery conversion.

Known methods of encoding video images on the basis of pulse code modulation, differential pulse code modulation, statistical coding and coding with prediction, see, for example, the book: Upref Digital image processing Part 2. - M.: Mir, 1982, s-688. These methods involve the coding of images with elementwise processing when a continuous signal is converted into a sequence of quantized samples, and then submitted to the code words in the form of zeros and ones.

Also known coding techniques based on the transformation, see, for example, the book: Nahmed. Crra Orthogonal transform in the processing of digital signals. - M.: Radio and communication, 1980, p.á192-201, including the execution of three operations: first, the image is subjected to two-dimensional orthogonal transformation, the resulting conversion factors quanthouse and further encoded for transmission over the channel tie is.

The disadvantage of the above methods - analogues is relatively low transmission rate of messages at a given quality of their recovery.

The closest to the technical nature of the claimed method is a method of compression and recovery of voice messages, described in the patent of the Russian Federation No. 2152646 And IPC7G 10 L 3/02 2000

Known prototype method is that the pre is identical to the transmitting side and the receiving side generates a random square matrix of quantized discrete samples of size mxm elements, each element of which belongs to the range of the quantized discrete samples of the speech signal. Continuous speech signal discretizing and quantum discrete samples. Then form the matrix of normalized values of the speech signal of size NxN elements so that each element aj,iwhere j=1,2,...,N; i=1,2,...,N, assign quantized discrete value of the reference speech signal, the k-th number of which is determined in accordance with the expression: k=j+N·(i-1). The matrix of quantized samples of the speech signal of size NxN elements convert to digital mind by forming sets the zero and unit elements in the form of a rectangular matrix of size Nxm and mxN elements. To do this, generate a random rectangular matrix of the unit is different, and zero elements of size Nxm and mxN elements. Then transform of a random rectangular matrix of size Nxm and mxN elements by dividing the elements of each row of a random rectangular matrix of size Nxm elements on the amount of units of the corresponding row and dividing the elements of each column of a random rectangular matrix of size mxN elements on the amount of units of the corresponding column. Then calculate the resulting matrix of size NxN elements by successive multiplication obtained after the conversion of rectangular matrix of size Nxm elements, a random square matrix of quantized discrete samples of size mxm elements and obtained after the conversion of rectangular matrix of size mxN elements. Calculate the sum of the squared differences between the elements obtained by multiplying the matrix of size NxN elements and the elements of the matrix of normalized values of the speech signal of size NxN elements. Then invert each element of a random rectangular matrix of size Nxm and mxN elements, convert them and consistently Peremohy obtained after the conversion of rectangular matrix of size Nxm elements, a random square matrix of quantized discrete samples of size mxm elements and obtained after the conversion of rectangular matrix of size mxN elements. The calc is slayt the sum of the squared differences between the elements obtained by multiplying the output matrix of size NxN and the elements of the matrix of normalized values of the speech signal of size NxN elements. Subtract this amount from similar amounts received in the previous step, and, in the case of a positive difference, keep the inverted value of the item, and otherwise perform his repeated inversion. After that convey a lot of zero and unit elements of a rectangular matrix of size Nxm and mxN elements over the communication channel. Take lots of zero and unit elements of a rectangular matrix of size Nxm and mxN elements of the communication channel and convert by dividing the elements of each row of a rectangular matrix of size Nxm elements on the amount of units of the corresponding row and dividing the elements of each column of a rectangular matrix of size mxN elements on the amount of units of the corresponding column. Then form the matrix of the restored normalized signal values of size NxN elements by successive multiplication obtained after the conversion of rectangular matrix of size Nxm elements, a random square matrix of quantized discrete samples of size mxm elements and obtained after the conversion of rectangular matrix of size mxN elements. In the last step, the matrix is restored normalized values of the speech signal of size NxN elements transform in the continuous speech signal.

Prototype method allows, without compromising the quality of the restoration is the next, to increase the data transmission rate to the amount at which you can conduct telephone conversations via low-speed digital communication channels.

The disadvantage of this method the prototype is still relatively low data transmission rate. Because the method-prototype directly encode the quantized discrete timing messages, you can restore messages with a specified quality it is necessary to encode a large array of data, which limits the applicability of this method for transferring still images.

The aim of the invention is to develop a method of compression and recovery messages that help increase the speed of information transmission with maintaining a given quality of recovery messages.

This objective is achieved in that in the known method for compressing and restoring messages previously on transmitting and receiving sides is identical generate random square matrix of size mxm elements that represent digital information signal in the form of a matrix of normalized values, generate a random rectangular matrix of size Nxm and mxN elements, convert them by dividing the elements of each row of a random rectangular matrix of size Nxm elements on the amount of units of the corresponding row and divide elem is now each column of a random rectangular matrix of size mxN elements on the amount of units of the corresponding column, calculate the resulting matrix of size NxN elements, by successive multiplication of the transformed random rectangular matrix of size Nxm on a random square matrix of size mxm and transformed random rectangular matrix of size mxN elements, then compute the RMS error between the elements of the output matrix of size NxN elements and the elements of the matrix of normalized values of size NxN elements, and then invert each element of a random rectangular matrix of size Nxm and mxN elements, and after the inversion of each element in a random rectangular matrix transform it by dividing the elements of each row of a random rectangular matrix with inverse element of size Nxm elements on the amount of units of the corresponding row and dividing elements each column of a random rectangular matrix with inverse element of size mxN elements on the amount of units of the corresponding column, re-calculate the resulting matrix of size NxN elements, by successive multiplication of the transformed random rectangular matrix with the inverse element of size Nxm on a random square matrix of size mxm and transformed random rectangular matrix with the inverse element of size mxN, the Le which calculates a root mean square error between the elements recomputed the output matrix of size NxN elements with the elements of the matrix of normalized values of size NxN elements, transmit over the communication channel random rectangular matrix of size Nxm, take the link to this matrix and on the receiving end converts the received random rectangular matrix of size Nxm and mxN by dividing the elements of each line of a random rectangular matrix of size Nxm elements on the amount of units of the corresponding row and dividing the elements of each column of a random rectangular matrix of size mxN elements on the amount of units of the corresponding column, calculate the resulting matrix of size NxN elements, by successive multiplication of the transformed random rectangular matrix of size Nxm on a random square matrix of size mxm and transformed random rectangular matrix of size mxN, form the digital information signal, at the receiving and transmitting the parties generate k random key matrix of size Nxm and mxN elements. Each element of a random square matrix of size mxm elements belongs to the range-500 to+500, as digital information signal is received by k-matrices of the quantized samples fixed grayscale image dimensions MHM, where k>1. Generating at the receiving and transmitting sides identical normalization matrix of size NxN elements, the elements C(i,j) is calculated by the formulaDG is i=1, 2,...,N, j=1, 2,...,N. as a random rectangular matrix of size mxN elements take on a transmitting side transposed random rectangular matrix of size Nxm elements. Form k-matrices of coefficients of two-dimensional discrete cosine transform size MHM elements by successive multiplication of the matrix of the discrete cosine transform size MHM elements for each matrix quantized samples fixed grayscale video size MHM elements and the transposed matrix of the discrete cosine transform size MHM elements. Next, form the k-matrix of coefficients of two-dimensional discrete cosine transform of size NxN elements, according to the formula Ag(i,j)=Lg(i,j), where i=1, 2,...,N, j=1, 2,...,N, g=1,2,...,k, Lg(i,j) - i,j-th element of g-th matrix of the coefficients of the two-dimensional discrete cosine transform size MHM elements, Ag(i,j) - i,j-th element of g-th matrix of the coefficients of the two-dimensional discrete cosine transform of size NxN elements, and choose N<M. Then form k-matrices of the normalized values of size NxN elements, by multiplying each coefficient of the matrix of coefficients of two-dimensional discrete cosine transform of size NxN elements of Ag(i,j) to the corresponding element normalization m is tricy size NxN elements. Each of the key matrices of size Nxm and mxN elements sum modulo 2, respectively, with direct and transposed random rectangular matrix of size Nxm and mxN elements, and after calculating the root mean square error between the corresponding elements of each of the output matrix of size NxN elements and the matrix of normalized values of size NxN elements, calculate their total amount. After inverting each element of a random rectangular matrix of size Nxm resulting sum is compared with the previous total. The communication channel transmit a random rectangular matrix of size Nxm elements, and at the receiving side after multiplying k random matrices of size Nxm on random matrix of size mxm and k-random matrices of size mxN, convert the resulting matrix of size NxN elements by element-by-element division of their elements to the corresponding elements of the normalization matrix of size NxN elements. The obtained k-matrix recovered coefficients of size NxN elements complement with zeros to the size of the MHM elements. Restore k-matrices a fixed grayscale images by successive multiplication of the transposed matrix of discrete cosine transform size MHM items on the k-matrices restored coefficients TLD the cluster discrete-cosine transform size MHM elements and a matrix of discrete cosine transform size MHM elements. For the formation of k-matrices of the quantized samples fixed grayscale video size MHM elements every element of Sg(x,y), where x=1,2,...,M; y=1,2,...,M, g=1,2,...,k, assigned a quantized value of the corresponding pixel of the k-fixed grayscale images of size NxN. To represent the k-matrices of the quantized samples fixed grayscale video size MHM elements in the form of k-fixed grayscale images, each pixel of the k-fixed grayscale video assign the value of the corresponding element of k-magic restored quantized samples fixed grayscale video size MHM elements.

Thanks to the new essential features by performing discrete cosine transform on the k-matrices of the quantized samples fixed grayscale video is go to view the video in the form of k-matrices of spectral coefficients. To reduce the digital representation of the video encode and transmit not all spectral coefficients, but only the N2spectral coefficients of the spectral region with a maximum energy.

In this way, all without compromising the quality of the message recovery, to increase the data transmission rate of up to ve who icine, where it is possible to transfer a fixed grayscale images in low-speed digital communication channels (32 kbps).

The analysis of the level of technology has allowed to establish that the analogues, characterized by a set of characteristics is identical for all features of the claimed technical solution is available, which indicates compliance of the claimed method the condition of patentability “novelty”. Search results known solutions in this and related areas of technology in order to identify characteristics that match the distinctive features of the prototype of the characteristics of the claimed method, showed that they do not follow explicitly from the prior art. The prior art also revealed no known effect provided the essential features of the claimed invention transformations on the achievement of the technical result. Therefore, the claimed invention meets the condition of patentability “inventive step”.

The claimed method is illustrated by drawings:

- Figa - random square matrix of size mxm elements.

- PIGB - variant random square matrix of size mxm elements.

- Figure 2 - formation of the k-matrix of coefficients of two-dimensional discrete cosine transform size MHM elements.

- 3 - the formation of k-matp the C coefficients of the two-dimensional discrete cosine transform of size NxN elements.

- Figa, 4B - forming normalization matrix of size NxN elements.

- 5 - the formation of k-matrices of the normalized values of size NxN elements.

- Figa - forming k-random key matrix of size Nxm elements.

- PIGB - option k is a random key matrix of size Nxm elements.

- FIGU - forming k-random key matrix of size mxN elements.

- Figg - option k is a random key matrix of size mxN elements.

- Fig 7a - forming k-random rectangular matrix of size Nxm elements.

- PIGB - option k is a random rectangular matrix of size Nxm elements.

- FIGU - forming k-random rectangular matrix of size Nxm elements.

- Figg - forming k-random rectangular matrices of size mxN elements.

- Figd transformation k is a random rectangular matrix of size Nxm elements.

- File transformation k is a random rectangular matrix of size mxN elements.

- Fig - forming k-resulting matrices of size NxN elements.

- Figa is a variant of the inverse element of a random rectangular matrix of size Nxm elements.

- Figure 10 is a transmission of a random rectangular matrix of size Nxm elements of a digital communication channel.

- Figa - variant matrix of coefficients of two-dimensional discrete cosine transform of asmerom elements.

- PIGB - the graph of the absolute values of the first row of the coefficient matrix of the two-dimensional discrete cosine transform size MHM elements.

- Fig - formation of k-matrices restored coefficients of two-dimensional discrete cosine transform of size NxN elements.

- Fig - formation of k-matrices restored coefficients of two-dimensional discrete cosine transform size MHM elements.

- Fig - formation of k-matrices restored quantized samples fixed grayscale image.

The possibility of implementing the inventive compression method and message recovery is explained by the following. If necessary, the transmission channel of communication messages, the volume of which exceeds the capacity of the communication channel or transmission which requires unacceptably large time interval, use various methods to reduce the volume of transmitted messages.

For example (see the book: Upref. Digital image processing. Part 1. - M.: Mir, 1982, pp.96-118), the encoded message is present in the form of the product matrix of the reference vectors in a matrix of coecients. For this purpose, use one of the known methods: the discrete cosine transform, fast Fourier transform, transformation of karunen-Loev, Wavelet transform, and others.

<> This technique causes some reduction in the amount of information required for transmission over the communication channel, and the simultaneous achievement of the required quality.

At the receiving end the received message restore.

Thus, when a deterioration in the quality of information transmitted reduce the amount of information required for transmission. At the same time, the volume of transmitted information about the matrix of support vectors and the matrix of coecients is still large, which does not meet the requirements of modern communication channels, while maintaining the required quality. In the proposed method solves the problem of reducing the amount of information while maintaining the required quality.

The proposed method involves carrying out the following actions.

The formation of the transmitting and receiving sides of a random square matrix of size mxm elements (hereinafter denote it as [B]mxm), each element of which belongs to the range-500 to+500 (see figa, 1B). The size m of the matrix [V]mxmchosen empirically based on the size of the message - M. Experimental studies show that the quality of approximation of the transmitted message size m is 1/5-1/4 of the size of the message. The operation of forming the matrix [B]mxmcan the be performed using a random numbers generator. To fulfill the requirements of the identity matrix [B]mxmthe receiver is similar to the matrix of the transmitter elements of the matrix [B]mxmcan be generated on the transmission side and transmitted over a digital communication channel at the receiving side, for example, as part of synchrophasing.

As k messages that are subject to compression and restoration, we consider the k-fixed grayscale video, which form a k-matrices of the quantized samples fixed grayscale video size MHM elements, for each of the video image by setting each element of Sg(x,y), where x=1,2,...,M; y=1,2,...,M, the quantized value of the corresponding pixel fixed grayscale video pictures, g=1,...,k the number fixed grayscale image (see figure 1 of Annex 1).

In order to reduce the amount of information transmitted over the communication channel, using the discrete cosine transform, is described, for example, in the book: Nahmad, Crra Orthogonal transform in the processing of digital signals. - M.: Communication, 1980, s-159.

The matrix of coefficients of two-dimensional discrete cosine transform size MHM elements are formed on the basis of the expression [Lg(x,y)]MxM=[G(x,y)]MxMx[Sg(x,y)]MxMx[G(x,y)]MxMwhere [Lg(x,y)]MxMmatrix coefficientsfor discrete cosine transform g th fixed grayscale video size MHM elements, [Sg(x,y)]MxMmatrix of quantized samples of the g-th fixed grayscale video size MHM elements [G(x,y)]MxMmatrix direct discrete cosine transform, [G(x,y)]MxM-matrix inverse discrete cosine transform (see figure 2).

The most informative from the point of view of recovery live video, are the coefficients of the two-dimensional discrete cosine transform with maximum energy, located in the upper left quadrant of the matrix of coefficients of two-dimensional discrete cosine transform size MHM elements (see figa). They emit, forming a matrix of coefficients of two-dimensional discrete cosine transform of size NxN elements (hereinafter denote it as), on the basis of the expression Ag(i,j)=Lg(i,j), where i=1, 2,...,N, j=1, 2,...,N, Lg(i,j) - i,j-th element of the matrix of coefficients of two-dimensional discrete cosine transform g th fixed grayscale video size MHM elements, Ag(i,j) - i,j-th element of the matrix of coefficients of two-dimensional discrete cosine transform g th fixed grayscale video size NxN elements, and choose N≤M (see figure 3).

The magnitude of the coefficients of two-dimensional discrete-cosine the aqueous conversion to a large extent depends on their sequence numbers, what can be determined from the graph (see figb), where x-axis is the sequence number of the coefficients of the first row of the coefficient matrix of the two-dimensional discrete cosine transform size MHM elements, and on the y - axis of their absolute values. In order to eliminate the dependence of the elements of the matrix of coefficients of two-dimensional discrete cosine transform of size NxN elements from their location in the matrix and in the future more accurately they can be approximated, it is necessary to perform the operation of rationing. The essence of this operation is that on the transmitting and receiving sides is identical normalization form a matrix of size NxN elements (hereinafter denote it as [s]NxN)whose elements C(i,j) is calculated by the formula(see figure 4)obtained experimentally. When this takes into account the peculiarity of the coefficient matrix of the two-dimensional discrete cosine transform size MHM elements, which consists in the arrangement of the coefficients with the maximum energy in the upper left quadrant and the dependence of coefficient values from their ordinal numbers (i and j).

Then form k-matrices of the normalized values of the two-dimensional discrete cosine transform of size NxN elements, by multiplying each element of Ag(i,j) matrix of the cylinder is having a two-dimensional discrete cosine transform g th fixed grayscale video size NxN elements on the corresponding element C(i,j) normalization matrix of size NxN elements (see 5). Similarly, using the normalization matrix, the normalized coefficients of the discrete cosine transform compression standard JPEG.

Further, similarly to the method prototype used an approach based on the representation matrix of the normalized coefficients of the two-dimensional discrete cosine transform of size NxN elements [V(k)]NxNas a product of three matrices: the converted rectangular matrix of size Nxm elements (hereinafter denote it as), a random square matrix of size mxm elements [B]mxmand converted to a rectangular matrix of size mxN elements (hereinafter denote it as(see Fig). Then, when encoding matrix [V(k)]NxNon the transmitting end must find the optimal matrixandwhich upon multiplication with the matrix [B]mxmform the resulting matrix of size NxN elements (hereinafter denote this matrix as)that is closest according to the specified criteria to the matrix [V(k)]NxN.

Matrixandform a by product of a random rectangular matrix [E]Nxmat the k-th random key matrix [YCL (k)]Nxmand transposed random rectangular matrix [E]

T
Nxm
at the k-th random key matrix [XCL(k)]mxNaccordingly, where the matrix [YCL(k)]Nxmand [XCL(k)]mxNare the key k-th fixed grayscale video size mxN (see figv, 7G).

Feature matrixandis that they can be easily converted to digital form. This is achieved by the fact that the elements of these matrices has the following limitations:

- matrix elementsandaccept values in the range from zero to one;

- non-zero elements in each row of the matrixequal in amount to form a unit;

- non-zero elements of each column of the matrixequal in amount to form a unit.

Under such restrictions, if the elements of each row of the matrixmultiply by the number of nonzero elements in this row, it will obtain the matrix [Y(k)]Nxmwhose elements are defined only on the set of “1” and “0”. Similarly, if the elements of the each column of the matrix multiply by the number of nonzero elements in the column, it will obtain the matrixthe elements of which are defined only on the set of “1” and “0”.

The procedure that implements a search on the transmission side of the optimal matricesanddescribed in detail in the method-prototype (see RF patent №2152646 And IPC7G 10 L 3/02 2000), the difference lies in the fact that the inversion is performed only in the matrix [E]Nxm(see Fig.9). When inverting element in the matrix [E]Nxmsimultaneously invert the elements of random rectangular matrices [Y(k)]Nxmand [X(k)]mxN.

Thus, the representation matrix of the normalized values of the two-dimensional discrete cosine transform of size NxN elements [V(k)]NxNin digital form on the transmission side is carried out on the basis of generating a set of zero and unit elements in the form of a random rectangular matrix of size Nxm (matrix [E]Nxm(see figa, 7b) and k is a random key matrix of size mxN and Nxm elements (matrix [XCL(k)]mxNand [YCL(k)]Nxm(see figv, 6g). Then the data matrix transform by dividing the elements of each row of a random rectangular matrix of size Nxm elements on the amount of units of the corresponding row, i.e. its weight - vy(k) (see CED is from: American. Algebraic coding theory - M.: Mir, 1971, p.12) (see Figg) and dividing the elements of each column of a random rectangular matrix of size mxN elements on the amount of units of the corresponding column, i.e., the weight of vx. Thereby form the matrixand(see file).

Similar to the prototype method calculates the resulting matrix of size NxN elements, i.eby successive multiplication obtained after the conversion of rectangular matrix of size Nxmrandom square matrix of size mxm elements [B]mxmand obtained from the conversion of rectangular matrix of size mxN(see Fig).

Matrixshould be closest to the matrix [V(k)]NxNaccording to a certain criterion. It is known (see, for example, the book: Upref. Digital image processing. Part 1. - M.: Mir, 1982, s-127)that one of the major objective criteria proximity is the standard error. Minimizing the root mean squared error, achieve minimal differences between matrices [V(k)]NxNandTherefore, calculate the sum of squared differences between the elements of the resulting Matri is s of size NxN and the corresponding matrix elements of the normalized values of the two-dimensional discrete cosine transform of size NxN elements [V]NxN. Then invert each element of a random rectangular matrix of size Nxm elements (see Fig.9) and convert them in a similar manner as was described for the transformation matrix [Y(k)]Nxmand [X(k)]mxN(see figa, 7b, 7b, 7G, 7D, 7E). Consistently Peremohy obtained after conversion, k is a random rectangular matrix of size Nxm elements of a random square matrix of size mxm elements and obtained after conversion, k is a random rectangular matrix of size mxN elements.

Because, in the matrix [E]Nxmcontained inverted element, after transformation has led to changes in the values of matrix elements [Y(k)]Nxmand [X(k)]mxNand consequently led to changes in the matrixandchange the values of the elements of the output matrixThen, to evaluate the degree of approximation of the matrixto [V(k)]NxNre-count k-sums of squared differences between the elements k of the resulting matrices of size NxN elements and elements of the k-matrices of the normalized values of the two-dimensional discrete-casinosno the conversion of size NxN elements. Then summarize the obtained k-sum of squares of the difference and subtract the resulting sum of squares of a difference from similar amounts received in the previous step. In case of a positive difference, i.e. reduce the standard error, retain the inverted value of the item, and otherwise perform his repeated inversion.

Similarly produce the inversion of all the bits in the matrix [E]NxNand achieve the minimum mean-square error between matricesand [V(k)]NxNthat clearly indicates the optimality of the generated matrices [Y(k)]Nxm, [X(k)]mxNandi.e. achieving the best quality at a given fixed amount of transmitted information.

Transmit the set of zero and unit elements of a rectangular matrix [E]Nxmthe communication channel (see figure 10). In figure 10 character ⊕ denotes matrix multiplication.

At the receiving side is the channel of communication many of the zero and unit elements of a rectangular matrix [E]Nxm. Then calculate the matrix [Y(k)]Nxmand [X(k)]mxNby works of the matrix [E]Nxmthe matrix [YCL(k)]Nxmand works transposed matrix [E]Nxmthe matrix [XCL(k)]mxNrespectively. Then convert the form of the matrix [Y(k)] Nxmand [X(k)]mxNby dividing the elements of each row of a rectangular matrix of size Nxm elements on the amount of units of the corresponding row, i.e. the weight of vy(see figa) and dividing the elements of each column of a rectangular matrix of size mxN elements on the amount of units of the corresponding column, i.e., the weight of vx(see Figg). Thus, at the receiving side form the matrixand

Form the matrix of the restored normalized values of the two-dimensional discrete cosine transform of size NxNby successive multiplication of the k-random rectangular matricesrandom square matrix [V]mxmand k-random rectangular matrices(see Fig).

To obtain the recovered coefficients of real dimension should be done dekorirovaniya. Given that at the receiving side was formed normalization matrix [S]NxNk-matrices reconstructed values of the two-dimensional discrete cosine transform of size NxN elements (hereinafter denote it as) formed by dividing the value of each i,j-th element of the matrixto meet the third element of the normalization matrix of size NxN elements (see Fig).

To recover the transmitted message must form a k-matrices restored coefficients of two-dimensional discrete cosine transform size MHM elements (hereinafter denote it as). This operation is carried out by assigning each i,j-th element of the matrix of reconstructed coefficients of two-dimensional discrete cosine transform of size NxN elements, each i,j-th element of the matrix of reconstructed coefficients of two-dimensional discrete cosine transform size MHM elements, as well as other items write zeros (see Fig).

Next, form the k-matrices restored quantized samples fixed grayscale video by multiplying the transposed matrix of discrete cosine transform size MHM items on the k-matrices restored coefficients of two-dimensional discrete cosine transform size MHM elements and a matrix of discrete cosine transform size MHM elements (see Fig), i.e. on the basis of the formula:wherematrix restored quantized samples of the g-th fixed grayscale video size MHM elements.

At the last stage are the matrix in a still grayscale of the video image by setting each pixel to a fixed grayscale video the value of the corresponding element of the matrix

To assess the possibility of achieving the formulated technical result when using the inventive method of compression and message recovery was conducted simulation on the PC. The size of the random square matrix [B]mxmwas 128x128 elements. This matrix size [In]mxmchosen on the assumption that the original message used fixed grayscale image of size 512x512 pixels. In the proposed method, a high compression ratio of the original message is achieved due to the fact that formation at the receiving side k-mentioned fixed grayscale video images in a digital communication channel, you need to pass the number of binary units, defined by the dimensions of the matrix [E]Nxm. In the General case, the matrix [E]Nxmare rectangular. But during simulation N is taken equal to 16, a m=128. This value of N due to the requirements to the quality of the reconstructed video. Empirical studies show that when leaving 1/16 spectral coefficients of two-dimensional discrete cosine transform of l is of the first upper quadrant of the matrix of coefficients of two-dimensional discrete cosine transform of size 512x512 elements peak signal-to-noise ratio for the original and the restored video is about 30 dB.

The achievable compression ratio can be found by the formula:

Figure 8 in the numerator of the specified formula says that for coding directly fixed grayscale of the video image, i.e. the value of each pixel is in the range 0-255, requires 8 bits. When choosing N=16, m=128 and M=512, the resulting compression ratio amounted to 16 times. When k=2, i.e. when passing through one of the matrix [E]Nxmtwo grayscale images total compression was 32 times. When using the prototype method for compressing messages resulting ratio was 16 times at peak signal-to-noise ratio of the order of 14 dB. Objective assessment of the quality of the restored using the inventive method, the video shows that the peak signal-to-noise ratio for the original and the restored video is 28.7 dB. Received the restored video is depicted in figure 2 of Annex 1.

The analysis of computational complexity have shown that the complexity of the proposed procedure coding/decoding is approximately proportional to the value of m2. Therefore, the proposed method for the compression and recovery of speech can be implemented on modern processors signal processing.

1. The compression method and message recovery, which consists in the fact that preliminary on transmitting and receiving sides is identical generate random square matrix of size m× m elements represent digital information signal in the form of a matrix of normalized values, generate a random rectangular matrix of size N×m m×N elements, convert them by dividing the elements of each row of a random rectangular matrix of size N×m items in the amount of units of the corresponding row and dividing the elements of each column of a random rectangular matrix of size m×N elements on the amount of units of the corresponding column, calculate the resulting matrix of size N×N elements by successive multiplication of the transformed random rectangular matrix of size N×m on a random square matrix of size m×m and the transformed random rectangular matrix of size m×N elements, then compute the RMS error between the elements of the output matrix of size N×N elements and the elements of the matrix of normalized values of size N×N elements, and then invert each element of random rectangular matrices of size N×m m×N elements and after the inversion of each element in a random rectangular matrix transform it by dividing the elements of each row of the random rectangular matrix with inverse element of size N×m items in the amount of units corresponding to the her line and dividing the elements of each column of a random rectangular matrix with inverse element size m× N elements on the amount of units of the corresponding column, re-calculate the resulting matrix of size N×N elements by successive multiplication of the transformed random rectangular matrix with the inverse element of size N×m on a random square matrix of size m×m and the transformed random rectangular matrix with the inverse element of size m×N, and then compute the RMS error between the elements recomputed the output matrix of size N×N elements with the elements of the matrix of normalized values of size N×N elements pass over the communication channel is a random rectangular matrix of size n×m, take the link to this matrix and on the receiving end converts the received random rectangular matrix of size N×m m×N by dividing the elements of each line of a random rectangular matrix of size N×m items in the amount of units of the corresponding row and dividing the elements of each column of a random rectangular matrix of size m×N elements on the amount of units of the corresponding column, calculate rezultirase matrix of size N×N elements, by successive multiplication of the transformed random rectangular matrix of size N×m on a random square matrix of size mm and converted to a random rectangular matrix of size m× N, form a digital information signal, characterized in that each element of a random square matrix of size mxm elements belongs to the range -500÷+500, as digital information signal is received by k matrices of the quantized samples fixed grayscale image sizes M×M, where k>1, the receiving and transmitting sides additionally generate k random key matrix of size N×m m×N elements, normalization matrix of size N×N items, where C(i,j) is calculated by the formula

where i=1, 2,..., N, j=1, 2,..., N,

and as a random rectangular matrix of size mxN elements take on a transmitting side transposed random rectangular matrix of size Nxm elements, form k matrices of coefficients of two-dimensional discrete cosine transform of size M×M elements by successive multiplication of the matrix of the discrete cosine transform of size M×M elements in each matrix of the quantized samples fixed grayscale video size M×M elements and the transposed matrix of the discrete cosine transform of size M×M elements of the form k matrices of coefficients of two-dimensional discrete cosine transformation is s of size N× N elements, according to the formula Ag(i,j)=Lg(i,j), where i=1,2,..., N,j=1,2,..., N,g=1,2,...,k, Lg(i,j) - i,j-th element of the 0-th coefficient matrix of two-dimensional discrete cosine transform of size M×M elements, Ag(i,j) - i-j-th element of g-th matrix of the coefficients of the two-dimensional discrete cosine transform of size N×N elements, and choose N<M, then form k matrix of normalized values of size N×N elements by multiplying each coefficient of the matrix of coefficients of two-dimensional discrete cosine transform of size N×N elements Ag(i,j) to the corresponding element of the normalization matrix of size N×N elements, then each of the key matrices of dimensions N×m m×N elements sum modulo 2, respectively, with direct and transposed random rectangular matrix of dimensions N×m and mxN elements, and after calculating the root mean square error between the corresponding elements of each of the output matrix of size N×N elements and a matrix of normalized values of size N×N elements calculate their total amount, after inverting each element of a random rectangular matrix of size N×m the resulting sum is compared with the previous total, and the communication channel transmit a random rectangular matrix of size Nxm elements, and the and the receiving side after multiplying k random matrices with dimensions N× m by a random matrix of size m×m and k random matrices with dimensions m×N convert the resulting matrix sizes N×N elements by element-by-element division of their elements to the corresponding elements of the normalization matrix of size N×N items, the k matrix recovered coefficients of size N×N elements complement with zeros to the size of a M×M elements, then restore the k matrices fixed grayscale images by successive multiplication of the transposed matrix of the discrete cosine transform of size M×M elements by k matrices restored coefficients of two-dimensional discrete cosine transform of size M×M elements and a matrix of discrete cosine transform of size M×M elements.

2. The method according to claim 1, characterized in that for the formation of k matrices of the quantized samples fixed grayscale video size M×M elements, each element Sg(x,y), where x=1,2,...,M; y=1,2,...,M, g=1,2,...,k, assigned a quantized value of the corresponding pixel k fixed grayscale images of size N×N.

3. The method according to claim 1, characterized in that for representation of k matrices of the quantized samples fixed grayscale video size M×M elements in the form of k NEPAD the author grayscale images, each pixel k fixed grayscale video assign the value of the corresponding element of k matrices restored quantized samples fixed grayscale video size M×M elements.



 

Same patents:

The invention relates to a digital camcorder, and more particularly to a device is a digital camcorder that uses methods of video compression that is compatible with the decoding device MPEG-2

The invention relates to filter the data, in particular to the way the signal adaptive filtering for reducing blocking effect and noise trim, signal adaptive filter and a machine-readable medium for storing program

The invention relates to techniques for television, in particular to the means of decoding the image signal

The invention relates to television technology, in particular to the transmission of video signals over a narrow-band channels, and for encoding broadband signals to reduce their bandwidth when the loss of information does not distort the overall perception

FIELD: electrical communications; data processing including reduction of data redundancy.

SUBSTANCE: proposed process includes similar way of generation of random quadrature matrix measuring m x m items and k random key matrices measuring N x m and m x N items on sending and receiving ends. Then k matrices of quantum readings of motionless gray-level video picture measuring M x M items are formed from k motionless gray-level video pictures which are then converted into product of three following matrices: random rectangular matrix measuring N x m items, random square matrix measuring m x m items, and random rectangular matrix measuring m x N items; in the process items of rectangular matrix measuring N x m items are transferred to communication channel. On receiving end k matrices of recovered quantum readings of motionless gray-level video pictures measuring M x M items are formed around random matrix measuring N x m items received from communication channel, as well as around random quadrature matrix measuring m x m items, and random rectangular matrix measuring m x N items, and motionless gray-level video pictures are produced from mentioned k matrices of recovered quantum readings.

EFFECT: enhanced data transfer speed at desired quality of recovered messages.

4 cl, 24 dwg

FIELD: technologies for data filtering.

SUBSTANCE: when a frame is formed of blocks of preset size, following operations are performed: generation of blocking information for decrease of blocking effect and contouring information for decrease of contouring noise from coefficients of preset pixels of upper and left limiting areas of data block, when a frame, received by decomposition of image data in a stream of binary bits for inverse quantizing, is an inner frame, and adaptive filtering of image data, passing through inverse quantizing and inverse discontinuous cosine transformation, in accordance to generated information of blocking and information of contouring. That is why blocking effect and contouring noise can be removed from an image, restored from image on basis of blocks, to improve the image, restored from compression.

EFFECT: decreased blocking effect and contouring noise.

2 cl, 7 dwg

FIELD: data filtration technologies, in particular, signaling adaptive filtration for lower blocking effect and contour noise.

SUBSTANCE: during forming of frame, following operations are performed: production of blocking information for decreasing blocking noise and production of contouring information for decreasing contouring noise of coefficients of previously given pixels of upper and left threshold areas of data block, when frame, received by decomposition of image data in the stream of binary digits for inverse quantizing is an internal frame, and adaptive filtration of image data passing through inverse quantizing and inverse discontinuous cosine transformation, in accordance to produced blocking information and contouring information. Thus, blocking effect and contouring noise can be removed from image, restored from image on basis of blocks, improving the image restored from compression.

EFFECT: decreased blocking effect and contouring noise of encoding with high compression level.

2 cl, 7 dwg

FIELD: data filtration technologies, in particular, signaling adaptive filtration for lower blocking effect and contour noise.

SUBSTANCE: during forming of frame of blocks of given size, following operations are performed: production of blocking information for decreasing blocking noise and production of contouring information for decreasing contour noise of coefficients of previously given pixels of upper and left threshold areas of data block, when frame, received by decomposition of image data in the stream of binary digits for inverse quantizing is an internal frame, and adaptive filtration of image data passing through inverse quantizing and inverse discontinuous cosine transformation, in accordance to produced blocking information and contouring information. Thus, blocking effect and contouring noise can be removed from image, restored from image on basis of blocks, improving the image restored from compression.

EFFECT: decreased blocking effect and contouring noise of encoding with high compression level.

2 cl, 7 dwg

FIELD: technology for encoding multimedia objects.

SUBSTANCE: method for encoding a multimedia object includes following stages: multimedia object is encoded for producing a bit stream and information about quality is added to bit stream, while information about quality denotes quality of multimedia object relatively to given position or relatively to given part of bit stream, while information about quality is provided in quality tags, aforementioned quality tag provides a values of quality tag, and value of quality tag characterizes distortion in encoded multimedia object being reproduced, when bit stream is truncated in point, related to quality tag.

EFFECT: development of improved and efficient method/system for encoding multimedia objects.

13 cl, 2 dwg

FIELD: electrical communications; data digital computation and processing including reduction of transferred information redundancy.

SUBSTANCE: proposed message compression and recovery method includes pre-generation of random quadrature matrix measuring m x m constituents and k random key matrices measuring N x m and m x N constituents on transmitting and receiving ends, and generation of quantum reading matrix of fixed half-tone video pattern measuring M x M constituents. Matrices obtained are transformed to digital form basing on addition and averaging of A images, each image being presented in the form of product of three matrices, that is, two random rectangular matrices measuring N x m and m x N constituents and one random quadrature matrix measuring m x m constituents. Transferred to communication channel are constituents of rectangular matrix measuring N x m constituents. Matrix of recovered quantum readings of fixed half-tone video pattern measuring M x M constituents is generated basing on rectangular matrix measuring N x m constituents received from communication channel as well as on random quadrature matrix measuring m x m constituents and random rectangular matrix of m x N constituents, and is used to shape fixed half-tone video pattern.

EFFECT: enhanced error resistance in digital communication channel during message compression and recovery.

2 cl, 26 dwg, 1 app

FIELD: video communications, in particular, technology for masking decoder errors.

SUBSTANCE: in accordance to one variant of invention, system and method decode, order and pack video information to video data packets for transfer via communication line with commutated channels, due to which system conceals errors, caused by loss of video data packets, when system receives, unpacks, orders and decodes data packets. In accordance to another variant, system and method decode and pack video information so that adjacent macro-blocks may not be positioned in same data packets. Also, system and method may provide information, accompanying packets of video data for simplification of decoding process. Advantage of described scheme is that errors caused due to data loss are distributed spatially across whole video frame. Therefore, areas of data, surrounding lost macro-blocks, are decoded successfully, and decoder may predict movement vectors and spatial content with high degree of precision.

EFFECT: improved quality of image.

4 cl, 10 dwg

FIELD: method for decreasing visual distortions in frame of digital video signal, which is encoded in blocks and then decoded.

SUBSTANCE: block type is determined in accumulator to encoding method for block, selected in accordance to given set of encoding type. For achieving technical result, i.e. decreasing visual distortions caused by limit of block, filtration is performed in the method, which is carried out depending on frame blocks types around the limit of block.

EFFECT: decreased visual distortions, increased reliability and efficiency.

9 cl, 6 dwg, 2 tbl

FIELD: radio engineering, possible use for digital processing of video signals, transferring the image.

SUBSTANCE: in accordance to the invention, the image being processed is divided on blocks with following transformation of each block using discontinuous quantum transformation, result coefficients are quantized and encoded, supporting points are computed and linear interpolation is performed, while before the stage of supporting point selection, one of the supporting points on edge limit of block is selected and a supporting point on opposite limit block is calculated using additional low frequency filters, after that linear interpolation is performed between thus computed supporting points.

EFFECT: improved quality of compressed video image with insignificant CPU resource costs.

2 cl, 4 dwg

FIELD: engineering of systems for encoding digital video signals, in particular, indication of values of quantization parameters in video encoding system.

SUBSTANCE: method and device for encoding a digital video series are claimed, where indication of quantization parameter is given out in encoded bit stream for use during decoding. Indication of information concerning the quantization parameter is ensured by insertion of SQP value - series level quantization parameter value. In particular, instead of encoding absolute values of parameters of quantization of image/section, indication of difference ΔQP between series level quantization parameter SQP and QP of image/section, is given out.

EFFECT: increased efficiency when encoding digital video signals and reduced speed of data transmission in bits.

4 cl, 8 dwg

Up!