Decoding method based on two-dimensional image data read by area-scan imaging mobile platform

Abstract

A decoding method based on two-dimensional image data read by an area-scan imaging mobile platform. The decoding method is an optical-storage signal processing method that is based on a Viterbi algorithm and is applicable to two-dimensional frame data, and conveniently implements the reading of two-dimensional data by means of a two-dimensional data matrix corresponding to two-dimensional image frames acquired by an area-scan imaging mobile platform. During the process of decoding, the effect of a decoded row on data being decoded can be eliminated on the basis of a two-dimensional partial response matrix, thereby solving the problem of inter-symbol interference and inter-track interference occurring between read signals and a convolutional effect being thus generated due to the length of a read focused light spot being much greater than a minimum marker signal length during reading on a high-density information surface. Information data recorded by an image acquired by an area-scan imaging technology is recovered with a high level of accuracy, thereby improving the recovery accuracy of two-dimensional frame information data under interference and noise conditions.

Description

A decoding method for two-dimensional image data read based on a surface imaging mobile platform Technical Field

[0001] The present invention relates to the technical field of optical storage signal processing, and in particular to a decoding method for two-dimensional image data read based on a surface imaging mobile platform. Background Art

[0002] The rapid development of big data and artificial intelligence technologies has led to traditional magnetic-based big data storage technologies facing severe bottlenecks in energy consumption, capacity, and lifespan. The evolution of optical disc recording formats from the original CD to the BD has utilized shorter laser wavelengths and higher NA focusing lenses to continuously reduce the focused beam spot diameter, increasing optical resolution while simultaneously reducing the disc's track pitch and minimum mark signal length, thereby improving recording density. Traditional optical information signal processing technology has traditionally employed serial readout of single-channel recorded data, with isolation tracks between each channel. Decoding is performed line by line on a single channel basis. The corresponding PRML technology primarily addresses intersymbol interference (ISI) to achieve accurate decoding. However, with BD (25GB per disc), this technology approach has reached its limits, as serial readout limits read speeds.

[0003] To achieve high-density and high-speed reading of larger amounts of information data, in addition to shortening the distance between symbols, it's also necessary to reduce the channel spacing. For example, data can be recorded using a dot matrix and read at high speed using a parallel readout method. For stable and accurate decoding, surface imaging technology is used to capture an image of the recording medium, extract the image data, and obtain two-dimensional frame data. However, when reading high-density information surfaces, the focused light spot is much larger than the shortest mark signal length, resulting in a two-dimensional signal with greater interference. Furthermore, the computational complexity of the two-dimensional PRML algorithm is the square of the one-dimensional computational complexity, making decoding more difficult and resulting in low information data recovery accuracy.

[0004] Summary of the Invention

[0005] The present invention provides a decoding method for two-dimensional image data read based on a surface imaging mobile platform to overcome the above technical problems.

[0006] In order to achieve the above object, the technical solution of the present invention is:

[0007] A method for decoding two-dimensional image data read by a surface imaging mobile platform comprises the following steps:

[0008] S1: Acquire a two-dimensional image of recorded data in a recording medium having recorded data by a surface imaging mobile platform to obtain a Page image corresponding to a recording page in a recording format of the recording medium; the surface imaging mobile platform includes a surface moving platform and a surface imaging device;

[0009] S2: Preprocessing the Page image to obtain a two-dimensional data matrix based on pixel values ​​of the image, thereby obtaining a two-dimensional data matrix after matrix reduction processing;

[0010] The number of elements in the rows and columns of the two-dimensional data matrix after the matrix reduction process is equal to the number of recording dots in the rows and columns of the recording medium;

[0011] S3: Based on the two-dimensional data matrix after the shrinkage process, the initial parameter matrix W with C rows and D columns is set C×D , get the optimal parameter matrix W' C×D , to update the pixel values ​​of the elements in the two-dimensional data matrix after the matrix reduction processing, and obtain an updated two-dimensional data matrix after the matrix reduction processing;

[0012] S4: Select the two-dimensional partial response matrix PR with k1 rows and k2 columns k , based on the Viterbi algorithm, the first Decode the pixel values ​​of the elements of the row; where k1≥2, k2≥2;

[0013] S5: The row number of the two-dimensional data matrix after the updated shrinkage process When the decoded The pixel values ​​of the elements of the row and the two-dimensional partial response matrix PR k , obtain the final decoded value of the element in the tth row and qth column of the updated two-dimensional data matrix SZ' after the shrinking process; in the order of raster scanning, sequentially decode the elements in the tth row and qth column of the updated two-dimensional data matrix after the shrinking process. The pixel values ​​of the elements in the row are decoded to complete the decoding of the recorded data in the recording medium; wherein T is the number of recording dots in the row of the recording medium.

[0014] Beneficial effects: The present invention is a decoding method for two-dimensional image data read based on a surface imaging mobile platform. It is an optical storage signal processing method suitable for two-dimensional frame data based on the Viterbi algorithm. The two-dimensional data matrix corresponding to the two-dimensional image frame obtained by the surface imaging mobile platform conveniently realizes the two-dimensional data reading. In the decoding process, it can eliminate the influence of the decoded rows on the data being decoded based on the two-dimensional partial response matrix, and solves the problem of inter-code crosstalk and inter-track crosstalk between the read signals and the convolution effect caused by the focused light spot read being much larger than the shortest mark signal length when reading on a high-density information surface. The information data recorded in the image obtained by the surface imaging technology is restored with high precision, and the accuracy of the recovery of the two-dimensional frame information data under the conditions of crosstalk and noise is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following is a brief introduction to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0016] FIG1 is a flow chart of a data decoding method according to the present invention;

[0017] FIG2 is a schematic diagram of a data recording format suitable for two-dimensional image frame data in an embodiment of the present invention;

[0018] FIG3 is a schematic diagram of a marking bit region when calculating an offset in an embodiment of the present invention;

[0019] FIG4 is a schematic diagram of a correction process for a two-dimensional data matrix in an embodiment of the present invention;

[0020] FIG5 is a schematic diagram of a rotation offset process of a two-dimensional data matrix in an embodiment of the present invention;

[0021] FIG6 is a schematic diagram of a two-dimensional data matrix shrinkage process according to an embodiment of the present invention;

[0022] FIG7 is a state transition diagram of a novel two-dimensional Viterbi algorithm in an embodiment of the present invention;

[0023] FIG8 is a schematic diagram of a decoding process of two-dimensional image frame data in an embodiment of the present invention;

[0024] FIG9 is a schematic diagram of a surface imaging mobile platform in an embodiment of the present invention. DETAILED DESCRIPTION

[0025] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0026] This embodiment provides a method for decoding two-dimensional image data read by a surface imaging mobile platform, which is characterized by comprising the following steps, as shown in FIG1 :

[0027] S1: Acquire a two-dimensional image of the recorded data in a recording medium having the recorded data by a surface imaging mobile platform to obtain a Page image corresponding to a recording page in a recording format of the recording medium; the surface imaging mobile platform includes a surface moving platform and a surface imaging device, as shown in FIG9 ;

[0028] Preferably, the recording format of the recording medium includes a plurality of recording media sheets arranged in sequence according to a raster scan order; the recording media sheets include a plurality of recording data blocks arranged in sequence according to a raster scan order; the recording data blocks include a plurality of recording pages arranged in sequence according to a raster scan order; the recording pages include a plurality of recording dots arranged in sequence according to a raster scan order;

[0029] A Page separation band is provided between two adjacent recording pages; a mark bit area and an address bit area corresponding to the mark bit area are provided inside the Page separation band;

[0030] Specifically, as shown in FIG2 ; in this embodiment, the recording format of the recording medium having recorded data is a three-level structure,

[0031] The first level is the recording media sheet, each recording media sheet contains several recording data blocks, the second level is the recording block, each recording data block contains several recording pages, the third level is the recording page, each recording page contains several recording dots, each recording dot records 1 recording data, and the recording data is not limited to binary, and can be multi-base data as needed.

[0032] Specifically, in this embodiment, each record page (Page) contains md rows and nd columns of dots. Two adjacent record pages (Page) are separated by a Page separation band consisting of fixed values ​​(e.g., 0) in xdv rows and xdh columns, respectively. The separation band contains a mark bit region and an address bit region. The mark bit region consists of l*w dots in l rows and w columns; the address bit region consists of r*s dots in r rows and s columns.

[0033] Specifically, the recording data block Block includes a Page with mp rows and np columns, and two adjacent blocks are separated by a block separation band composed of fixed values ​​(for example, 0) in xpv rows and xph columns, wherein xpv and xph are greater than or equal to NPR, wherein NPR is the number of PR response coefficients, and the boundary of each block forms a frame with a fixed response value (for example, 0); the recording media sheet includes a Block with ms rows and ns columns.

[0034] Preferably, the method for obtaining a Page image in a recording medium having recorded data is as follows:

[0035] S11: fixing the recording medium on the surface moving platform of the surface imaging moving platform, and photographing the recording medium by the surface imaging device of the surface imaging moving platform to obtain an initial photographed image;

[0036] Specifically, in an embodiment of the present invention, in the initial captured image obtained by the surface imaging device of the surface imaging mobile platform, there will be multiple complete recording pages. Furthermore, in the initial captured image, there will be multiple mark positions corresponding to the complete recording pages. Since the relative position of the mark position relative to the corresponding recording page is determined in the recording medium with recorded data, the corresponding recording page can be accurately located through the mark position area.

[0037] S12: According to the initial captured image, obtain the number of pixels a along the X-axis direction and the number of pixels b along the Y-axis direction of the marking area in the initial captured image; obtain the pixel matrix B of the marking area a×b ;

[0038] S13: Based on the pixel matrix B of the marking area a×b , obtaining the defocus amount of the surface imaging device of the surface imaging mobile platform to obtain the final position coordinates of the surface mobile platform on the Z axis; and moving the surface mobile platform along the Z axis;

[0039] Specifically, in an embodiment of the present invention, a three-dimensional coordinate system is established using the surface imaging mobile platform, wherein the surface imaging mobile platform is capable of translation along the X-axis, Y-axis, and Z-axis, and is parallel to the XY plane of the three-dimensional coordinate system. Furthermore, in this embodiment, assuming that a recording medium is fixed to the surface imaging mobile platform, the boundaries of the pages on the recording medium are parallel to the X-axis and Y-axis, respectively.

[0040] Specifically, this embodiment can obtain the number of pixels a along the x-axis and the number of pixels b along the y-axis in the marker area and the pixel value of each pixel through the existing template matching algorithm; and then obtain the pixel matrix B of the marker area. a×b ; Select the pixel matrix B of the marking area within the pixel matrix of the marking area a×b The two-dimensional defocus matrix M of the center point pixel coincidence am×bm ; Select pixel matrix B in the area of ​​the marker outside the matrix a×b The two-dimensional difference matrix N of the center point pixel coincidence an×bn , and respectively calculate the defocus amount of the surface imaging device of the surface imaging mobile platform, the row offset and the vertical offset of the marking area.

[0041] Preferably, the method for determining the final position coordinates of the surface moving platform on the Z axis is as follows:

[0042] S131: Selecting a two-dimensional defocus matrix M am×bm , the two-dimensional defocus matrix M am×bm The pixel matrix B of the marked area a×b The center point pixels coincide with each other, and am <a,bm<b;

[0043] Among them, am is the two-dimensional defocus matrix M am×bm The number of pixels along the X-axis; bm is the two-dimensional defocus matrix M am×bm The number of pixels along the Y axis;

[0044] S132: Obtain the coordinate of the surface moving platform on the Z axis as Z n When the two-dimensional defocus matrix M am×bm The pixel values ​​of all elements in are summed to obtain the defocus amount of the surface imaging device of the surface imaging mobile platform, that is, the maximum pixel value; wherein n represents the coordinate number of the surface mobile platform on the Z axis;

[0045] Specifically, the total number of coordinates of the surface moving platform on the Z axis depends on the moving range of the surface moving platform on the Z axis and the step size set each time the platform moves.

[0046] Preferably, the defocus amount of the surface imaging device of the surface imaging mobile platform is obtained as follows:

[0047] LJ=max(SUM n )

[0048] Where: SUM n The coordinate of the surface moving platform on the Z axis is Z n The two-dimensional defocus matrix M when am×bm The sum of the pixel values ​​of all elements in ; LJ is the defocus value of the surface imaging device of the surface imaging mobile platform; m i,j is the two-dimensional defocus matrix M am×bm The element in row i and column j of the _{\mathbf {{ ...

[0049] S133: Obtain the two-dimensional defocus matrix M according to the defocus amount of the surface imaging device of the surface imaging mobile platform. am×bm When the sum of the pixel values ​​of all elements in is the maximum pixel value, the final position coordinate of the surface moving platform on the Z axis;

[0050] S134: Move the surface moving platform along the Z axis to the final position coordinates of the surface moving platform on the Z axis.

[0051] S14: Based on the pixel matrix B of the marking area a×b , obtaining the row offset and vertical offset of the marking area to move the surface moving platform;

[0052] Preferably, the method for moving the surface moving platform is as follows:

[0053] S141: Select the two-dimensional difference matrix N an×bn , the two-dimensional difference matrix N an×bn The pixel matrix B of the marked area a×b The center point pixels coincide with each other, and an>a,bn>b; where an is the two-dimensional differential matrix N an×bn The number of pixels along the Y axis; bn is the two-dimensional difference matrix N an×bn The number of pixels along the X-axis;

[0054] S142: According to the two-dimensional differential matrix Na n×bn , obtain the first row offset matrix XL, the second row offset matrix XR, the first vertical offset matrix YU and the second vertical offset matrix YD; as shown in Figure 3;

[0055] Preferably, the first row offset matrix XL, the second row offset matrix XR, the first vertical offset matrix YU and the second vertical offset matrix YD are obtained as follows:

[0056] The first offset matrix XL is obtained as follows:

[0057] The second row offset matrix XR is obtained as follows:

[0058] And xr=xl

[0059] The first vertical offset matrix YU is obtained as follows:

[0060] The second vertical offset matrix YD is obtained as follows:

[0061] And yd=yu

[0062] Where: n 2,1 Represents the two-dimensional difference matrix N an×bn The element in row 2 and column 1; n 2,xl Represents the two-dimensional difference matrix N an×bn xl is the number of pixels in the first row offset matrix XL along the X-axis; xr is the number of pixels in the first row offset matrix XR along the X-axis, where xr = xl; yu is the number of pixels in the first vertical offset matrix YU along the Y-axis; yd is the number of pixels in the second vertical offset matrix YD along the Y-axis, where yu = yd;

[0063] S143: respectively obtain the sum SUMXL of the pixel values ​​of all elements in the first row offset matrix XL, the sum SUMXR of the pixel values ​​of all elements in the second row offset matrix XR, the sum SUMYU of the pixel values ​​of all elements in the first vertical offset matrix YU, and the sum SUMYD of the pixel values ​​of all elements in the second vertical offset matrix YD;

[0064] S144: Moving the surface moving platform, the method is as follows:

[0065] When SUMXL-SUMXR>0, move the surface moving platform in the negative direction of the X axis until the value of SUMXL-SUMXR is less than the set row offset threshold;

[0066] When SUMXL-SUMXR<0, move the surface moving platform in the positive direction of the X axis until the value of SUMXL-SUMXR is less than the set row offset threshold;

[0067] When SUMYU-SUMYD>0, move the surface moving platform in the positive direction of the Y axis until SUMYU-SUMYD is less than the set vertical offset threshold;

[0068] When SUMYU-SUMYD<0, move the surface moving platform in the negative direction of the Y axis until SUMYU-SUMYD is less than the set vertical offset threshold.

[0069] S15: According to the moved surface moving platform, a Page image corresponding to the mark bit area and the recording page Page in the recording format is acquired.

[0070] Specifically, in this embodiment, after S11 - S15 are executed, one Page image is obtained; by repeatedly executing S11 - S15 , multiple Page images can be obtained, and eventually all Page images in the recording medium can be obtained.

[0071] S2: Preprocessing the Page image to obtain a two-dimensional data matrix based on the pixel values ​​of the image; obtaining a two-dimensional data matrix after matrix reduction processing; the number of elements in the rows and columns of the two-dimensional data matrix after matrix reduction processing is equal to the number of recording dots in the rows and columns of the recording medium;

[0072] Specifically, according to the Page image, a two-dimensional data matrix of pixel values ​​in the image can be obtained based on existing methods;

[0073] S21: performing a rotation operation on the two-dimensional data matrix, as shown in FIG5 , so that the recording points dot in the Page image are parallel to the X axis and the Y axis respectively, to obtain the two-dimensional data matrix after the rotation operation;

[0074] Specifically, the method for rotating a two-dimensional data matrix is ​​a prior art and will not be described in detail here.

[0075] S22: performing a correction operation on the two-dimensional data matrix after the rotation operation, as shown in FIG4 , so that the number of elements in the rows and columns of the two-dimensional data matrix after the rotation operation is equal to the number of elements in the set rows and the number of elements in the set columns, thereby obtaining the two-dimensional data matrix after the correction operation;

[0076] Specifically, due to factors such as recording error, shooting error, and Page image interception error, the page image actually obtained is unlikely to be of the ideal size. In an ideal state, a single Page image should have P ip Row P jp A single recording dot in the recording medium occupies p in the imaging photograph. x Line q xThe ideal size of the two-dimensional data matrix is ​​P ip ×p x Row and P jp ×q x Therefore, a correction operation needs to be performed on the two-dimensional data matrix after the rotation operation to achieve the desired size. Specifically, the method for performing the correction operation on the two-dimensional data matrix after the rotation operation is to resample the two-dimensional data matrix after the rotation operation by row and column respectively. The method for performing the correction operation on the two-dimensional data matrix after the rotation operation in this embodiment is prior art and is only used here, so it will not be described in detail.

[0077] S23: For the two-dimensional data matrix after the correction operation, subtract the average value of the pixel values ​​of the area where the data is not recorded in the recording medium from the pixel value of each element in the matrix to obtain a two-dimensional data matrix after the DC component is removed;

[0078] S24: fusing the two-dimensional data matrix after removing the DC component according to the address bits to obtain a fused two-dimensional data matrix;

[0079] Specifically, according to the order of address bit records in the page, multiple two-dimensional data matrices after the DC component is removed are spliced, and the spliced ​​two-dimensional data matrix after the DC component is removed is the fused two-dimensional data matrix.

[0080] S25: performing a shrinking process on the fused two-dimensional data matrix so that the number of elements in the rows and columns of the two-dimensional data matrix after the shrinking process is equal to the number of recording dots in the rows and columns of the recording medium, and obtaining the two-dimensional data matrix after the shrinking process.

[0081] Specifically, as shown in FIG6 , in the embodiment of the present invention, it is assumed that there is J in the recording medium. h Line J l The elements in the fused two-dimensional data matrix are divided into J according to the number of elements. h Line J l The process involves creating a matrix block with columns; each matrix block corresponds to an element at a corresponding position in the two-dimensional data matrix after the shrinking process. The sum of the pixel values ​​of all pixels in each matrix block is then taken as the element in the corresponding two-dimensional data matrix after the shrinking process. This allows us to obtain the value of each element in the two-dimensional data matrix after the shrinking process, resulting in the final two-dimensional data matrix after the shrinking process.

[0082] S3: Based on the initial parameter matrix W with C rows and D columns C×D , according to the two-dimensional data matrix after the shrinkage processing, the parameter matrix update method is used to obtain the optimized parameter matrix W' C×D, to update the element values ​​in the two-dimensional data matrix after the matrix reduction processing, and obtain an updated two-dimensional data matrix after the matrix reduction processing;

[0083] Preferably, the method for obtaining the updated two-dimensional data matrix after the matrix reduction process is as follows:

[0084] S31: Set the initial parameter matrix W with C rows and D columns C×D ,

[0085] The initial parameter matrix W C×D satisfy:

[0086] in, Represents the initial parameter matrix W C×D The Row, No. The value of the element in the column is also the element in the center of the initial parameter matrix; << means much less than the symbol; w c×1 Represents the initial parameter matrix W C×D The value of the element in column 1; w 1×d Represents the initial parameter matrix W C×D The value of the element in row 1; w c×D Represents the initial parameter matrix W C×D The value of the element in column D in ; w C×d Represents the initial parameter matrix W C×D The value of the element in the Cth row; c represents the initial parameter matrix W C×D The row number in ; d represents the initial parameter matrix W C×D The number of the column in ; Indicates rounding up operation;

[0087] Specifically, set the parameter matrix W C×D , W C×D is a two-dimensional data matrix with large values ​​at the center and small values ​​at the edges;

[0088] S32: Based on the two-dimensional data matrix after the shrinkage process and the initial parameter matrix W C×D ,Based on the parameter matrix updating method, obtain the optimized parameter matrix W';

[0089] Preferably, the parameter matrix updating method is as follows:

[0090] S321: The first row and first column element SZ in the two-dimensional data matrix SZ after the matrix reduction process is used 1,1 As the center, select the reference matrix block CSZ with the same size as the initial parameter matrix 1,1 ;

[0091] Specifically, in this embodiment, at the positions of the elements in the selected reference matrix block that do not belong to the two-dimensional data matrix SZ after the matrix reduction process, the corresponding element values ​​are set to 0.

[0092] S322: Obtain the CSZ matrix according to the reference matrix block and the initial parameter matrix. 1,1 The corresponding intermediate calculation matrix

[0093] In the formula: CSZ 1,1 Represents the element SZ in the first row and first column of the two-dimensional data matrix SZ after the shrinking process 1，1 A reference matrix block centered on and of the same size as the initial parameter matrix; For CSZ 1,1 The corresponding intermediate calculation matrix;-

[0094] S323: Obtain the sum of the values ​​of all elements in the intermediate calculation matrix:

[0095] Where: Indicates CSZ 1,1 The corresponding intermediate calculation matrix The value of the element in row c and column d;

[0096] S324: Obtain updated parameter matrix W1;

[0097] W1=W+α×CSZ 1,1 ×(E(SZ 1,1 )-y)

[0098] Where: W1 is the updated parameter matrix; α is the empirical coefficient; E(SZ 1,1 ) is the first row and first column element SZ in the two-dimensional data matrix SZ after the matrix reduction process 1,1 The expected value of the pixel value;

[0099] S325: If E(SZ 1,1 )-y> the set convergence threshold, then in the order of raster scanning, the elements in the two-dimensional data matrix SZ after the shrinkage process are executed in steps S321-S324 until E(SZ t,q )-y≤the set convergence threshold, the updated parameter matrix obtained at this time is the optimized parameter matrix W' C×D ; E(SZ t,q ) represents the element SZ in the tth row and qth column of the two-dimensional data matrix SZ after the matrix reduction process. t,q The expected value of the pixel value.

[0100] Specifically, in an embodiment of the present invention, starting from the first row and first column element of the two-dimensional data matrix SZ after the matrix shrinkage processing, the processing is performed in sequence according to the raster scanning order, and a reference matrix block with the same size as the initial parameter matrix is ​​selected and multiplied with the initial parameter matrix to obtain an intermediate calculation matrix, and then the sum of the values ​​of all elements in the intermediate calculation matrix is ​​obtained to update the initial parameter matrix.

[0101] S33: Taking the element in the tth row and qth column of the two-dimensional data matrix SZ after the matrix reduction process as the center, obtain a reference matrix block CSZ with the same size as the initial parameter matrix t,q ;

[0102] Wherein, t is the row number of the two-dimensional data matrix SZ after the shrinking process, and is also the row number of the two-dimensional data matrix after the shrinking process, t=1,…,T; q is the column number of the two-dimensional data matrix SZ after the shrinking process, and is also the column number of the two-dimensional data matrix after the shrinking process, q=1,…,Q; T is the number of recording dots in a row of the recording medium, that is, the number of recording dots in a row of the two-dimensional data matrix SZ after the shrinking process; Q is the number of recording dots in a column of the recording medium, that is, the number of recording dots in a column of the two-dimensional data matrix SZ after the shrinking process;

[0103] S34: According to the optimization parameter matrix W' C×D , get with CSZ t,q The corresponding intermediate calculation matrix

[0104] In the formula: CSZ t,q represents a reference matrix block centered at the element in the t-th row and q-th column of the two-dimensional data matrix SZ after the matrix shrinkage processing and having the same size as the optimization parameter matrix; For CSZ t,q The corresponding intermediate calculation matrix;-

[0105] S35: Get with CSZ t,q The sum of the pixel values ​​of all elements in the corresponding intermediate calculation matrix; updating the value of the t-th row and q-th column of the two-dimensional data matrix SZ after the shrinkage processing to obtain an updated two-dimensional data matrix SZ' after the shrinkage processing; and updating the optimization parameter matrix based on the parameter matrix updating method.

[0106] Where: Indicates CSZ t,q The corresponding intermediate calculation matrix The value of the element in row c and column d in ; y t,qRepresents the value of the t-th row and q-th column of the updated two-dimensional data matrix after the shrinkage processing;

[0107] Updated two-dimensional data matrix after shrinkage processing

[0108] S4: Select the two-dimensional partial response matrix PR with k1 rows and k2 columns k , based on the Viterbi algorithm, the first Decode the pixel values ​​of the elements of the row;

[0109] Specifically, the two-dimensional partial response matrix PR is selected k , PR k It is a two-dimensional data matrix with size of k1 rows and k2 columns (k1, k2 ≥ 2). Among them, the two-dimensional partial response matrix is ​​a set value, which can be adjusted according to the optical characteristics of the recording spot and the pixel situation of the recording mark in the photograph. k The values ​​in can be symmetrically distributed or asymmetrically distributed according to the optical characteristics of the recorded light spot and the photos taken. Based on the Viterbi algorithm, the first Decoding the pixel values ​​of the elements of a row is a use of existing technology and will not be described in detail here.

[0110] S5: The row number of the two-dimensional data matrix after the updated shrinkage process When the decoded The elements of the rows and the two-dimensional partial response matrix PR k , obtain the final decoded value of the element in the tth row and qth column of the updated two-dimensional data matrix SZ' after the shrinking process; in the order of raster scanning, sequentially decode the elements in the tth row and qth column of the updated two-dimensional data matrix after the shrinking process. The pixel values ​​of the elements in the row are decoded to complete the decoding of the recorded data in the recording medium; wherein T is the number of recording dots in the column of the recording medium.

[0111] Preferably, the method for obtaining the final decoded value of the element in the t-th row and q-th column in the updated two-dimensional data matrix SZ' after the matrix reduction process is as follows:

[0112] S51: The two-dimensional partial response matrix PR k No. The element value of the row is the same as the updated two-dimensional data matrix SZ' after the shrinkage process. Multiply the decoded values ​​of the elements of the row to obtain the decoding intermediate matrix J';

[0113] Where: is the two-dimensional partial response matrix PR k The first row element in ; is the two-dimensional partial response matrix PR k The row elements; is the first in the updated two-dimensional data matrix SZ' after the shrinkage process Row, No. The decoded value of the column; is the t-1th row in the updated two-dimensional data matrix SZ' after the shrinkage process, The decoded value of the column; Indicates floor operation.

[0114] S52: Obtain the sum SUMJ' of all element values ​​in the decoding intermediate matrix J':

[0115] S53: Obtain the final decoding value of the element in the t-th row and q-th column in the updated two-dimensional data matrix SZ' after the matrix reduction process, which is J t,q -SUMJ';

[0116] Among them, J t,q is the pixel value of the element in the t-th row and q-th column in the updated two-dimensional data matrix SZ' after the shrinkage processing;

[0117] S54: Based on the Viterbi algorithm, decode the final to-be-decoded value of the element in the t-th row and q-th column in the updated two-dimensional data matrix SZ' after the shrinkage processing, and obtain the final decoded value of the element in the t-th row and q-th column in the updated two-dimensional data matrix SZ' after the shrinkage processing.

[0118] Specifically, in this embodiment, the first The pixel values ​​of the rows are decoded row by row and point by point. First, the front The influence value of the row on the point to be decoded is the sum of all element values ​​in the decoding intermediate matrix J', SUMJ', and then subtract it from the value to be decoded to obtain the final decoded value. Effect on decoded values.

[0119] Specifically, a two-dimensional partial response matrix model PR is made k , PR k It is a two-dimensional matrix of size k1*k2 (k1, k2 ≥ 2). The size of the partial response model matrix can theoretically be any number. For the sake of convenience, the following example uses k=3 as an example.

[0120] In this embodiment, the matrix data is a symmetrical structure. In practical applications, the data in the matrix may also be an asymmetrical data structure according to the optical characteristics of the recorded light spot and the photographs taken.

[0121] If the usual Viterbi algorithm is used, PR k There are 2 possible migration states k1*k2 (k1, k2 ≥ 2) possibilities, in order to simplify the algorithm, eliminate PR k The influence of the first n rows (n≧1) in the model, at this time the number of possible migration states is reduced to 2(kn)*(kn)(n≧1), which greatly reduces the amount of calculation.

[0122] For example, set k = 3, n = 1, according to Create a state transition diagram, where a1, b1, c1, and d1 are all binary numbers, and the total number of possible transition states is 2. 4 =16, each state is as follows

[0123] Down:

[0124] Each state has four possible transition states. For example, the current state is The next state may be One of them is to draw a two-dimensional matrix model PR based on this state transition relationship. k The state transition diagram of the two-dimensional Viterbi algorithm. When k = 3, the state transition diagram is drawn as shown in Figure 7.

[0125] As shown in Figure 8, parameter matrix learning is performed first. The initial parameter matrix W is set. The feature block set containing the mark bit in the block to be decoded is taken and divided into two-dimensional reference matrix blocks by row. The reference matrix block is convolved with the initial parameter matrix, and the result is compared with the ideal data to obtain error data. The error data is input into the error function, and the parameter matrix is ​​updated. The above learning process is continued, and the updated parameter matrix is ​​used in the convolution operation until the optimal parameter matrix W' is obtained.

[0126] New 2D Viterbi decoding: Perform new 2D Viterbi decoding on the updated 2D data matrix after the matrix shrinkage process. Before decoding, crosstalk elimination is required, i.e., eliminating the previous OK( ) on the crosstalk effect of the current row, and then divide the updated two-dimensional data matrix after the shrinkage processing into a sequence of two-dimensional data matrices of size k1*k2 (k1, k2≥2) by row. For each element in the sequence, combined with the state transition diagram of the new two-dimensional Viterbi algorithm, it is compared with each state to obtain the most matching combination and obtain the decoded data.

[0127] The present embodiment is a decoding method for two-dimensional image data read by a surface imaging mobile platform. It is an optical storage signal processing method suitable for two-dimensional frame data based on the Viterbi algorithm. The two-dimensional data matrix corresponding to the two-dimensional image frame obtained by the surface imaging mobile platform conveniently realizes the two-dimensional data reading. During the decoding process, the influence of the decoded rows on the data being decoded can be eliminated based on the two-dimensional partial response matrix, so that the decoding algorithm can be greatly simplified. It solves the problem that when reading high-density information surfaces, because the focused light spot read is much larger than the shortest mark signal length, the read signals will have inter-code crosstalk and inter-track crosstalk, resulting in a convolution effect. The information data recorded by the image obtained by the surface imaging technology is restored with high precision, and the accuracy of the recovery of the two-dimensional frame information data under the conditions of crosstalk and noise is improved. The recorded information data is restored with high precision, solving the problem of improving the accuracy of the recovery of the two-dimensional frame information data under the conditions of crosstalk and noise.

[0128] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A decoding method for two-dimensional image data read by a surface imaging mobile platform, characterized in that, The steps are as follows: S1: Obtain a two-dimensional image of the recorded data in the recording medium with recorded data through a surface imaging mobile platform to obtain a Page image corresponding to the record page Page in the recording format of the recording medium; the surface imaging mobile platform includes a surface mobile platform and a surface imaging device; S2: Preprocess the Page image to obtain a two-dimensional data matrix based on the pixel values of the image to obtain a two-dimensional data matrix after reduction matrix processing; The number of elements in the rows and columns of the two-dimensional data matrix after reduction matrix processing is equal to the number of recording dots dot in the rows and columns of the recording medium; S3: According to the two-dimensional data matrix and parameter matrix update method after matrix reduction processing, based on the initial parameter matrix W with C rows and D columns set C×D , obtain the optimized parameter matrix W' C×D , so as to update the pixel values of the elements in the two-dimensional data matrix after matrix reduction processing, and obtain the updated two-dimensional data matrix after matrix reduction processing; S4: Select a two-dimensional partial response matrix PR of k1 rows and k2 columns k , based on the Viterbi algorithm, for the Decode the pixel values of the elements in the row; where k1≥2, k2≥2; S5: When the row number of the two-dimensional data matrix after the updated reduced matrix processing When, based on the decoded The elements of the row and the two-dimensional partial response matrix PR k , obtain the finally decoded value of the element in the q-th column of the t-th row in the updated two-dimensional data matrix SZ' after the reduction matrix processing; in the order of raster scanning, successively for the Decode the pixel values of the elements in the row to complete the decoding of the recorded data in the recording medium; where T is the number of recording dots dot in the column of the recording medium.

2. The decoding method for two-dimensional image data read by a surface imaging mobile platform according to claim 1, characterized in that, The method for obtaining the finally decoded value of the element in the t-th row and q-th column of the updated two-dimensional data matrix SZ' after reduction matrix processing is as follows: S51: Take the two-dimensional partial response matrix PR k The The element value of the row and the Multiply the decoded values of the elements of the row to obtain the decoded intermediate matrix J'; Where: is the first row element of the two-dimensional partial response matrix PR k in; is the two-dimensional partial response matrix PR k in the Row element; For the Row, the Decoded value of the column; For the (t-1)-th row in the updated two-dimensional data matrix SZ' after the contraction process, the Decoded value of the column; S52: Obtain the sum SUMJ' of all element values in the decoded intermediate matrix J'; S53: Obtain that the final value to be decoded of the element in the t-th row and q-th column of the updated two-dimensional data matrix SZ' after the matrix reduction is J t,q -SUMJ'; Among them, J t,q is the pixel value of the element at the q-th column and the t-th row in the updated two-dimensional data matrix SZ' after the contraction process; S54: Based on the Viterbi algorithm, decode the finally to-be-decoded value of the element in the t-th row and q-th column of the updated two-dimensional data matrix SZ' after reduction matrix processing to obtain the finally decoded value of the element in the t-th row and q-th column of the updated two-dimensional data matrix SZ' after reduction matrix processing.

3. A decoding method for two-dimensional image data read by a surface imaging mobile platform according to claim 1, characterized in that The recording format of the recording medium includes several recording medium sheets sheet arranged in sequence according to the raster scan order; the recording medium sheet sheet includes several recording data blocks Block arranged in sequence according to the raster scan order; the recording data block Block includes several record pages Page arranged in sequence according to the raster scan order; the record page Page includes several recording dots dot arranged in sequence according to the raster scan order; A Page separation band is provided between two adjacent record pages Page; a marker bit and an address bit corresponding to the marker bit area are provided inside the Page separation band.

4. A decoding method for two-dimensional image data read by a surface imaging mobile platform according to claim 1, characterized in that, In S1, the method for obtaining the Page image in the recording medium with recorded data is as follows: S11: Fix the recording medium on the surface mobile platform of the surface imaging mobile platform, and photograph the recording medium through the surface imaging device of the surface imaging mobile platform to obtain an initial photographed image; S12: Obtain the number of pixel points a in the X-axis direction and the number of pixel points b in the Y-axis direction of the marked area in the initial captured image; to obtain the pixel point matrix B of the marked area a×b ; S13: According to the pixel point matrix B of the marked bit area a×b , obtain the defocus amount of the surface imaging device of the surface imaging mobile platform, so as to obtain the final position coordinates of the surface mobile platform on the Z axis; and move the surface mobile platform along the Z axis direction; S14: According to the pixel point matrix B of the marker bit area a×b , obtain the row offset and vertical offset of the marker bit area to move the surface moving platform; S15: According to the moved surface mobile platform, obtain a Page image corresponding to the record page Page in the recording format corresponding to the marker bit area.

5. A decoding method for two-dimensional image data read by a surface imaging mobile platform according to claim 4, characterized in that In S13, the defocus amount of the surface imaging device of the surface imaging mobile platform is obtained as follows: LJ = max(SUM n ) Where: SUM n is the coordinate of the surface moving platform on the Z-axis as Z n when the sum of the pixel values of all elements in the two-dimensional defocus matrix M am×bm ; LJ is the defocus amount of the surface imaging device of the surface imaging moving platform; m i,j is the element in the i-th row and j-th column of the two-dimensional defocus matrix M am×bm , and n is the number of the coordinate of the surface moving platform on the Z-axis.

6. A decoding method for two-dimensional image data read by a surface imaging mobile platform according to claim 5, characterized in that, In S13, the method for obtaining the final position coordinate of the surface mobile platform on the Z axis is as follows: S131: Select the two-dimensional defocus amount matrix M am×bm , where the two-dimensional defocus amount matrix M am×bm coincides with the central pixel of the pixel point matrix B of the marked bit area a×b , and am < a, bm < b; where am is the two-dimensional defocus amount matrix M am×bm The number of pixel points in the X-axis direction; bm is the two-dimensional defocus amount matrix M am×bm The number of pixel points in the Y-axis direction; S132: Obtain the situation when the coordinate of the surface mobile platform on the Z-axis is Z n at which time, the two-dimensional Defocus amount matrix M am×bm Sum the pixel values of all elements in the matrix to obtain the defocus amount of the surface imaging device of the surface imaging mobile platform; where n represents the serial number of the coordinate of the surface mobile platform on the Z-axis. S133: Obtain the final position coordinates of the surface moving platform on the Z-axis when the sum of the pixel values of all elements in the two-dimensional defocus amount matrix M is the maximum pixel value according to the defocus amount of the surface imaging device of the surface imaging moving platform am×bm ; S134: Move the surface mobile platform along the Z axis to the final position coordinate of the surface mobile platform on the Z axis.

7. A decoding method for two-dimensional image data read by a surface imaging mobile platform according to claim 4, characterized in that, In S14, the method for moving the surface mobile platform is as follows: S141: Select the two-dimensional differential matrix N an×bn , the two-dimensional differential matrix N an×bn coincides with the central pixel of the pixel point matrix B a×b of the marked bit area, and an > a, bn > b; where an is the number of pixel points of the two-dimensional differential matrix N an×bn along the Y-axis direction; bn is the number of pixel points of the two-dimensional differential matrix N an×bn along the X-axis direction; S142: Obtain a first row offset matrix XL, a second row offset matrix XR, a first vertical offset matrix YU, and a second vertical offset matrix YD according to the two-dimensional differential matrix N an×bn , where the first row offset matrix XL, the second row offset matrix XR, the first vertical offset matrix YU, and the second vertical offset matrix YD are obtained; The first offset matrix XL is obtained as follows: The second row offset matrix XR is obtained as follows: and xr = xl The first vertical offset matrix YU is obtained as follows: The second longitudinal offset matrix YD is obtained as follows: and yd = yu where: n 2,1 represents the element in the 1st column of the 2nd row of the two-dimensional differential matrix N an×bn ; n 2,xl represents the element in the xl-th column of the 2nd row of the two-dimensional differential matrix N an×bn ; xl is the number of pixel points in the X-axis direction of the first row offset matrix XL; xr is the number of pixel points in the X-axis direction of the first row offset matrix XR, where xr = xl; yu is the number of pixel points in the Y-axis direction of the first vertical offset matrix YU; yd is the number of pixel points in the Y-axis direction of the second vertical offset matrix YD, where yu = yd S143: Obtain the sum SUMXL of the pixel values of all elements in the first row offset matrix XL, the sum SUMXR of the pixel values of all elements in the second row offset matrix XR, the sum SUMYU of the pixel values of all elements in the first vertical offset matrix YU, and the sum SUMYD of the pixel values of all elements in the second vertical offset matrix YD respectively; S144: Move the surface moving platform as follows: When SUMXL - SUMXR > 0, move the surface moving platform in the negative X-axis direction until the value of SUMXL - SUMXR is less than the set row offset threshold; When SUMXL - SUMXR < 0, move the surface moving platform in the positive X-axis direction until the value of SUMXL - SUMXR is less than the set row offset threshold; When SUMYU - SUMYD > 0, move the surface moving platform in the positive Y-axis direction until SUMYU - SUMYD is less than the set vertical offset threshold; When SUMYU - SUMYD < 0, move the surface moving platform in the negative Y-axis direction until SUMYU - SUMYD is less than the set vertical offset threshold.

8. A decoding method for two-dimensional image data read by a surface imaging mobile platform according to claim 1, characterized in that, In S2, the method for obtaining the two-dimensional data matrix after matrix reduction processing is as follows: S21: Perform a rotation operation on the two-dimensional data matrix so that the recording points dot in the Page image are respectively parallel to the X-axis and the Y-axis; to obtain the two-dimensional data matrix after the rotation operation; S22: Perform a correction operation on the two-dimensional data matrix after the rotation operation so that the number of elements in the rows and columns of the two-dimensional data matrix after the rotation operation is equal to the number of elements in the set rows and the number of elements in the set columns, and obtain the two-dimensional data matrix after the correction operation; S23: For the two-dimensional data matrix after the correction operation, subtract the average value of the pixel values of the area where no data is recorded in the recording medium from the pixel value of each element in the matrix to obtain the two-dimensional data matrix after removing the DC component; S24: According to the address bits, fuse the two-dimensional data matrix after removing the DC component to obtain the fused two-dimensional data matrix; S25: Perform a matrix reduction processing on the fused two-dimensional data matrix so that the number of elements in the rows and columns of the two-dimensional data matrix after the matrix reduction processing is equal to the number of recording points dot in the rows and columns of the recording medium, and obtain the two-dimensional data matrix after the matrix reduction processing.

9. A decoding method for two-dimensional image data read by a surface imaging mobile platform according to claim 1, characterized in that, In S3, the method for obtaining the updated two-dimensional data matrix after matrix reduction processing is as follows: S31: Set the initial parameter matrix W of C rows and D columns C×D , The initial parameter matrix W C×D satisfies: Among them, Denote the initial parameter matrix W C×D in the Row, the Element values of the column; << represents much less than symbol; w c×1 Represents the initial parameter matrix W C×D Value of the element in the first column; w 1×d Represents the initial parameter matrix W C×D Value of the element in the first row; w c×D Represents the initial parameter matrix W C×D Value of the element in the D-th column; w C×d Represents the initial parameter matrix W C×D Value of the element in the C-th row; c represents the initial parameter matrix W c×D Row number in; d represents the initial parameter matrix W C×D Column number in; Denote the ceiling integer operation; S32: According to the two-dimensional data matrix after the matrix reduction process and the initial parameter matrix W C×D , obtain an optimized parameter matrix W' based on the parameter matrix update method; S33: Taking the element at the t-th row and q-th column in the two-dimensional data matrix SZ after the matrix reduction processing as the center, a reference matrix block CSZ of the same size as the initial parameter matrix is obtained t,q ; S34: Obtain an intermediate calculation matrix corresponding to CSZ according to the optimized parameter matrix W' C×D t,q ​​ where: CSZ t,q represents a reference matrix block centered on the element in the t-th row and q-th column of the two-dimensional data matrix SZ after the reduction matrix processing and having the same size as the optimization parameter matrix; For the intermediate calculation matrix corresponding to CSZ t,q ; - S35: Obtain the sum of the pixel values of all elements in the intermediate calculation matrix corresponding to CSZ t,q ; update the value at the t-th row and q-th column of the two-dimensional data matrix SZ after the matrix reduction process to obtain the updated two-dimensional data matrix SZ' after the matrix reduction process; meanwhile, update the optimization parameter matrix based on the parameter matrix update method Wherein: Denote the intermediate calculation matrix corresponding to CSZ t,q ​ The value of the element in the c-th row and d-th column of; y t,q Indicates the value of the element in the t-th row and q-th column of the two-dimensional data matrix after the updated contraction matrix processing; Updated two-dimensional data matrix after contraction processing 10. A decoding method for two-dimensional image data read by a surface imaging mobile platform according to claim 1, characterized in that, In S3, the parameter matrix update method is as follows: S321: Using the element SZ at the first row and first column in the two-dimensional data matrix SZ after the matrix reduction process as the center, select a reference matrix block CSZ with the same size as the initial parameter matrix 1,1 ; 1,1 ; S322: Obtain an intermediate calculation matrix corresponding to CSZ according to a reference matrix block and an initial parameter matrix 1,1 corresponding intermediate calculation matrix where: CSZ 1,1 represents a reference matrix block centered on the element SZ at the first row and first column of the two-dimensional data matrix SZ after the reduction matrix processing, 1,1 and having the same size as the initial parameter matrix; For the intermediate calculation matrix corresponding to CSZ 1,1 ; - S323: Obtain the sum of the values of all elements in the intermediate calculation matrix: In the formula: Denote the intermediate calculation matrix corresponding to CSZ 1,1 ​ the value of the element in the c-th row and d-th column in; S324: Obtain the updated parameter matrix W1; W1 = W + α × CSZ 1,1 ×(E(SZ 1,1 ) - y) Where: W1 is the updated parameter matrix; α is the empirical coefficient; E(SZ 1,1 ) is the expected value of the pixel value of the element SZ 1,1 at the first row and first column in the two-dimensional data matrix SZ after matrix reduction processing; S325: If E(SZ 1,1 ) - y > the set convergence threshold, then, in the order of raster scanning, perform S321 - S324 on the elements in the two-dimensional data matrix SZ after the contraction process until E(SZ t,q ) - y ≤ the set convergence threshold. At this time, the updated parameter matrix obtained is the optimized parameter matrix W' C×D ; E(SZ t,q ) represents the expected value of the pixel value of the element SZ t,q at the t-th row and q-th column in the two-dimensional data matrix SZ after the contraction process.

Images