CN107065006B - Seismic signal coding method based on online dictionary updating - Google Patents

Abstract

A seismic signal coding method based on online dictionary updating belongs to a seismic signal data coding transmission method, solves the dictionary transmission problem caused by dictionary learning and sparse representation in seismic signal coding, and can be applied to various seismic signal measurement-based ground mapping. The invention comprises the following steps: (1) dividing input seismic signals into a plurality of groups according to a time sequence, and performing sparse coding on each group of data by using a dictionary in a cache to calculate a sparse coefficient; (2) quantizing and entropy coding the sparse coefficient in the step (1); (3) and reading the reconstructed data of the previous P transmitted groups from the cache, and performing dictionary learning by combining the sparse coefficients transmitted by the current group, thereby updating the dictionary required by sparse representation of the next group of data. According to the invention, by means of online dictionary updating, dictionary information does not need to be transmitted in real time under the premise of ensuring effective sparse representation of signals, so that data transmission data volume is effectively reduced, and the method can be suitable for various seismic signal high-speed acquisition application occasions.

Description

A Seismic Signal Coding Method Based on Online Dictionary Update

technical field

The invention belongs to a seismic signal data transmission method, in particular to a seismic signal lossy coding method.

Background technique

The surveying and mapping technology based on seismic signal measurement is one of the effective methods to measure the underground structure and mineral resources. In each survey, seismic signal measurement on the ground will generate more than 100T of data, and the bandwidth of current signal transmission is extremely limited, so it is necessary to reduce the amount of seismic signal data through seismic signal encoding technology before transmission. In the prior art, a seismic signal coding method based on discrete cosine transform is proposed, which can obtain a compression factor close to 3 times. There are also two-dimensional discrete cosine transform technology based on local seismic signal adaptation, so that the important features of the reconstructed seismic signal can be preserved. Furthermore, the seismic signal coding technology using adaptive wavelet packet can obtain higher compression factor and better reconstruction quality. Due to its better direction preservation characteristics, it is widely used in the feature extraction of seismic signals. The main idea of the above method is to use a suitable base or redundant dictionary to represent the seismic signal, so that the representation of the signal is sparse. In recent years, sparse representation through dictionary learning has received extensive attention, especially in image coding. It has been widely used in remote sensing images to learn dictionaries through double sparse models, thereby obtaining better coding results. These research results show the feasibility of applying dictionary learning and sparse representation in seismic signal coding.

Traditional coding methods based on dictionary learning and sparse representation often include the following two main methods: (1) sparse representation of online data acquired in real time through offline learned dictionaries. For this method, an offline training set needs to exist in advance, and the required dictionary information is obtained through the offline training set. Therefore, whether the sparse representation of online data is effective or not depends heavily on the correlation between offline data and online data. For actual seismic signal measurements, it is difficult to obtain a general offline training set applicable to different situations. (2) Use the real-time online data to train a dictionary, and use the dictionary to sparsely represent the real-time online data. In this method, it is necessary to transmit the dictionary, so that the sparse representation process at the codec side can be synchronized. Therefore, the transmission of dictionary information will increase the size of the encoded code stream, thereby reducing the encoding performance.

Invention content:

The invention proposes a seismic signal coding method based on online dictionary update, which belongs to the seismic signal data coding transmission method, solves the problem of how to transmit the dictionary caused by adopting dictionary learning and sparse representation in seismic signal coding, and can be applied to various seismic signals based on Underground mapping for signal measurements.

A seismic signal encoding method based on online dictionary update of the present invention includes an encoding step and a decoding step; wherein,

The encoding step includes:

Step 1. Divide the input seismic signals into multiple groups according to time sequence, and use the dictionary in the cache to perform sparse coding on each group of data, specifically:

Step 11: Divide the seismic signal data of the adjacent T tracks into one group, and process each group of data separately; Suppose that the current group of data is the Z group of data, which is represented as Y _z ; The data of each track is equally divided into several units, each unit _yi has a length of M×1, and _yi is sorted in columns; therefore, Y _z =[y ₁ ,...y _i ,...y _N ]; here it is assumed that each The length of the data recorded on the track is U, then there is the following relation: T×U=M×N;

Step 12: Read the dictionary D _z-1 in the cache. Given that the sparsity of the sparse coefficient matrix W _Z is L, optimize the solution to the following formula:

Step 2. Perform quantization and entropy coding on the sparse coefficients in step S1, specifically including:

Step 21: Quantize the sparse coefficient matrix using a uniform quantization method, as follows:

w _Z (i,j) represents the coefficient value with the coordinates (i, j) in the sparse coefficient matrix W _Z , Δ represents the quantization step size, Represents the quantization result of the coefficient value of (i, j), and round( ) represents the rounding operation;

Step 22. Create a non-zero coefficient position matrix PT consisting of a value of 0 and a value of 1, and the creation method is as follows:

Among them, abs( ) represents the absolute value operation;

Step 23, using arithmetic coding on the non-zero coefficient position matrix PT;

Step 24. For non-zero coefficients (corresponding to the position of PT(i,j)=1 Using Huffman coding;

Step 3. Read the reconstructed data of the previous P transmitted groups from the cache, and perform dictionary learning in combination with the sparse coefficients transmitted by the current group, thereby updating the dictionary required for the sparse representation of the next group of data, including:

Step 31. Calculate the reconstruction data of P+1 group p∈[ZP,Z] (the current group of data is the Z group), the calculation method is as follows:

in, unit in

Step 32: Solve the required dictionary D _Z according to the following optimization process:

in, a _i represents a constant describing the correlation between groups. Equation (2) is solved by iterative operation as follows:

Step 321, fix D _Z , W' can be calculated by the PS method described above;

Step 322, fix W', D _Z can be updated according to the MOD method:

Step 323, repeat the above-mentioned steps 321 and 322 to the specified number of iterations, and update the required dictionary D _Z ;

The decoding step includes:

Step 4. Perform inverse quantization and entropy decoding on the received sparse coefficients to generate a non-zero coefficient matrix W' _Z , as follows:

Step 41, performing Huffman decoding on the non-zero coefficient encoded code stream to obtain the non-zero coefficient w _c ;

Step 42: Perform inverse quantization on the non-zero coefficient w _c to obtain the inverse quantization coefficient w' _c , as follows:

w' _c =w _c ×Δ

Step 43, perform arithmetic decoding on the non-zero coefficient position matrix PT encoded code stream to obtain the non-zero coefficient position matrix PT, and combine the inverse quantization coefficients w' _c generated in step 42 to generate a non-zero coefficient matrix W'_Z;

Step 5. Perform seismic signal reconstruction, as follows:

Step 51: Read the dictionary D _z-1 in the cache, and generate the reconstructed signal Y' _z , as follows:

Y' _z =D _z-1 ×W' _Z

Step 52: Rearrange the reconstructed signal Y' _z =[y' ₁ ... y' _i ... y' _N ] (the length of each unit y' _i is M×1), and arrange several adjacent units Join the end and end together to form a trace in the way of columns, so the length of each trace is There are a total of T traces;

Step 6. Generate a dictionary D _z in the cache for the reconstruction of the next group of data, that is: read the reconstructed data of the previous P transmitted groups from the cache, and perform dictionary learning in combination with the sparse coefficients transmitted by the current group, so as to update The next set of dictionaries required for the sparse representation of the data include:

Step 61. Calculate the reconstruction data of P+1 groups p∈[ZP,Z] (the current group of data is the Z group), the calculation method is as follows:

in, unit in

Step 62: Solve the required dictionary D _Z according to the following optimization process:

in, a _i represents a constant describing the correlation between groups. Equation (2) is solved by iterative operation as follows:

Step 621, fix D _Z , W' can be calculated by the PS method described above;

Step 622, fix W', D _Z can be updated according to the MOD method:

Step 623: Repeat the above steps 321 and 322 to the specified number of iterations, and update the required dictionary D _Z .

The method of online dictionary updating does not require real-time transmission of dictionary information under the premise of ensuring effective sparse representation of signals, thereby effectively reducing the amount of data transmission data, and can be applied to various seismic signal high-speed acquisition applications.

Description of drawings

Fig. 1 is the flow chart of the present invention;

Fig. 2 is the partial signal in the test seismic signal data;

Figure 3 is the learned dictionary;

Figure 4 shows the results of the performance comparison of different methods.

Detailed ways

Example:

The present invention mainly includes:

S1: Divide the input seismic signal into multiple groups according to time sequence, and use the dictionary in the cache to sparsely encode each group of data;

S2: carry out quantization and entropy coding to the sparse coefficients in step S1;

S3: Read the reconstructed data of the previous P transmitted groups from the cache, and perform dictionary learning in combination with the sparse coefficients transmitted by the current group, so as to update the dictionary required for the sparse representation of the next group of data.

Further, step S1 is specifically:

S11: Divide the seismic signal data of the adjacent T traces into one group, and process each group of data separately; assuming that the current group of data is the Z group of data, which is represented as Y _z . The data of each trace is equally divided into several units, the length of each unit _yi is M×1, and the _yi is sorted by column. Therefore, Y _z =[y ₁ , . . . y _i , . . . y _N ]. Assuming that the length of data recorded on each track is U, there is the following relational formula: T×U=M×N.

S12: Read the dictionary D _z-1 in the cache, given that the sparsity of the sparse coefficient matrix W _Z is L, optimize the solution to the following formula:

For the solution of equation (1), we intend to use the PS method (“Partial search vector selection for sparse signal representation,” in NORSIG-03). The PS method is based on the OMP algorithm (orthogonal matching pursuit, "Comparison of basis selection methods," in Signals, Systems and Computers, 1996. Conference Record of the Thirtieth Asilomar Conference on), therefore, the flow of the OMP algorithm is first given:

The PS method modifies the search process of only the largest relevancy dictionary unit in the above step (1) to a search of several maximal relevancy dictionary units, thereby providing more search decisions to obtain better sparse vectors.

Step S2 is specifically:

S21: Quantize the sparse coefficient matrix by a uniform quantization method, as follows:

w _Z (i,j) represents the coefficient value with the coordinates (i, j) in the sparse coefficient matrix W _Z , Δ represents the quantization step size, Represents the quantization result of the coefficient value of (i,j), and round(·) represents the rounding operation.

S22: Create a non-zero coefficient matrix PT consisting of a value of 0 and a value of 1. The creation method is as follows:

Among them, abs( ) represents absolute value operation.

S23: Use arithmetic coding on the non-zero coefficient matrix PT.

S24: Huffman coding is used for the non-zero coefficients (w ^ri _,j corresponding to the position of PT(i,j)=1).

Step S3 is specifically:

S31: Calculate the reconstruction data of P+1 groups p∈[ZP,Z] (the current group of data is the Z group), the calculation method is as follows:

in, unit in

S32: Solve the required dictionary D _Z according to the following optimization process:

in, a _i represents a constant describing the correlation between groups.

Equation (2) can be solved iteratively through the following steps:

1) fixed D _Z , W' can be calculated by the PS method described above;

2) Fixed W', D _Z can be updated according to the MOD method ("Method of Optimal Directions for FrameDesign," in 1999 IEEE International Conference on Acoustics, Speech and SignalProcessing (ICASSP)):

3) Repeat the above steps 1) and 2) to the specified number of iterations to generate the required update dictionary D _Z .

Example 1:

1. The test seismic signal data comes from the UTAM image database (http://utam.gg.utah.edu/SeismicData/SeismicData.html), we choose the Find-Trapped-miners data as the test data, which contains 72 sensors, each Each sensor contains 135 traces;

2. Take 1600 time length samples for each trace, and the data of each 10 traces is one group, and some test data are shown in Figure 1;

1. Assuming that the current group is the third group (the first two groups of data have been encoded and the relevant data has been output to the decoding end and the cache), read the dictionary D2 in the cache for sparse encoding. _In this embodiment, the cache is The size of the dictionary is all 16 × 64, and the sparsity is 1/16 (the proportion of non-zero coefficients to the overall coefficients). Part of the data in W ₃ obtained by calculation is as follows;

0 0 0 0 0 0 0 0 0 0 0 0 0 191086.69 0 0 0 0 0 0 0 0 0 69516.63 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 263371.03 0 0 0 0 0 -275961.58 248225.21 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2. Quantize W ₃ , and the quantization step size is 1024; Huffman coding is used for the non-zero data after quantization, and arithmetic coding is used for the non-zero element position. We measure the difference between the original signal and the reconstructed signal by SNR. The SNR is 21.2dB.

3. By reading the first two sets of reconstruction data and W ₃ , the dictionary is updated, and the updated dictionary D ₃ is shown in FIG. 3 .

Example 2:

1. sparsely represent the fourth group of data with the dictionary D3 updated by the _third group of data, and update the dictionary D4 _;

2. The fifth group of data is sparsely represented by the dictionary D ₄ updated with the fourth group of data, and the dictionary D ₅ is updated;

3. Quantize and encode each group of data above, calculate the bit rate and measure the distortion through SNR;

4. Adjust different sparsity and repeat the above 3 steps to get the distortion at different bit rates.

5. In order to prove the effectiveness of the algorithm, we also conducted experiments on the seismic signal coding algorithm based on DCT, Curvelet and offline dictionary learning method (K-SVD+ORMP), and obtained the rate-distortion of different algorithms. The related experimental results are as follows: shown in Figure 4.

The specific embodiments described in this technology are merely illustrative of the spirit of the invention. Those skilled in the art to which the present invention pertains can make various modifications or additions to the described specific embodiments or substitute in similar manners, but will not deviate from the spirit of the present invention or go beyond the definitions of the appended claims range.

Claims (1)

1. a seismic signal encoding method based on online dictionary update, is characterized in that, comprises encoding step and decoding step; Wherein, The encoding step includes: Step 1. Divide the input seismic signals into multiple groups according to time sequence, and use the dictionary in the cache to perform sparse coding on each group of data, specifically: Step 11: Divide the seismic signal data of the adjacent T tracks into one group, and process each group of data separately; Suppose that the current group of data is the Z group of data, which is represented as Y _z ; The data of each track is equally divided into several units, each unit _yi has a length of M×1, and _yi is sorted in columns; therefore, Y _z =[y ₁ ,...y _i ,...y _N ]; here it is assumed that each The length of the data recorded on the track is U, then there is the following relation: T×U=M×N; Step 12: Read the dictionary D _z-1 in the cache. Given that the sparsity of the sparse coefficient matrix W _Z is L, optimize the solution to the following formula: Step 2. Perform quantization and entropy coding on the sparse coefficients in step S1, including: Step 21: Quantize the sparse coefficient matrix using a uniform quantization method, as follows: w _Z (i,j) represents the coefficient value with the coordinates (i, j) in the sparse coefficient matrix W _Z , Δ represents the quantization step size, Represents the quantization result of the coefficient value of (i, j), and round( ) represents the rounding operation; Step 22. Create a non-zero coefficient position matrix PT consisting of a numerical value of 0 and a numerical value of 1. The creation method is as follows: Among them, abs( ) represents the absolute value operation; Step 23, using arithmetic coding on the non-zero coefficient position matrix PT; Step 24: Huffman coding is used for the non-zero coefficients, and the non-zero coefficients correspond to the position of PT(i, j)=1. Step 3. Read the reconstructed data of the previous P transmitted groups from the cache, and perform dictionary learning in combination with the sparse coefficients transmitted by the current group, so as to update the dictionary required for the sparse representation of the next group of data, including: Step 31. Calculate the reconstruction data of P+1 group p∈[ZP,Z], the current group of data is the Z group, and the calculation method is as follows: in, unit in Step 32: Solve the required dictionary D _Z according to the following optimization process: in, a _i represents a constant describing the correlation between groups. Equation (2) is solved by iterative operation as follows: Step 321, fix D _Z , W' is calculated by PS method; Step 322, fix W', D _Z can be updated according to the MOD method: Step 323, repeat the above-mentioned steps 321 and 322 to the specified number of iterations, and update the required dictionary D _Z ; The decoding step includes: Step 4. Perform inverse quantization and entropy decoding on the received sparse coefficients to generate a non-zero coefficient matrix W' _Z , as follows: Step 41, performing Huffman decoding on the non-zero coefficient encoded code stream to obtain the non-zero coefficient w _c ; Step 42: Perform inverse quantization on the non-zero coefficient w _c to obtain the inverse quantization coefficient w' _c , as follows: w' _c =w _c ×Δ Step 43, perform arithmetic decoding on the non-zero coefficient position matrix PT encoded code stream to obtain the non-zero coefficient position matrix PT, and combine the inverse quantization coefficients w' _c generated in step 42 to generate a non-zero coefficient matrix W'_Z; Step 5. Perform seismic signal reconstruction, as follows: Step 51: Read the dictionary D _z-1 in the cache, and generate the reconstructed signal Y' _z , as follows: Y' _z =D _z-1 ×W' _Z Step 52: Rearrange the reconstructed signal Y' _z =[y' ₁ ... y' _i ... y' _N ], the length of each unit y' _i is M×1, and arrange adjacent units according to The columns are connected end to end to form a trace, so the length of each trace is There are a total of T traces; Step 6. Generate a dictionary D _z in the cache for the reconstruction of the next group of data, that is: read the reconstructed data of the previous P transmitted groups from the cache, and perform dictionary learning in combination with the sparse coefficients transmitted by the current group, so as to update The next set of dictionaries required for the sparse representation of the data include: Step 61. Calculate the reconstruction data of P+1 groups p∈[ZP,Z], the current group of data is the Z group, and the calculation method is as follows: in, unit in Step 62: Solve the required dictionary D _Z according to the following optimization process: in, a _i represents a constant describing the correlation between groups. Equation (2) is solved by iterative operation as follows: Step 621, fix D _Z , W' is calculated by PS method; Step 622, fix W', D _Z can be updated according to the MOD method: Step 623: Repeat the above steps 321 and 322 to the specified number of iterations, and update the required dictionary D _Z .