Digital audio watermarking method and device, electronic equipment and storage medium

Abstract

The invention provides a digital audio watermarking method and device, electronic equipment and a storage medium, and the method comprises the steps: carrying out the segmentation processing and transform domain processing of first audio information, and determining a watermark data embedding domain in each audio segment; performing digital watermark information embedding processing on the audio data in each watermark data embedding domain, determining a plurality of audio segments embedded with watermarks, and combining the audio segments to determine second audio information; performing segmentation processing and characteristic value statistical processing on the second audio information based on the candidate position and the candidate segment length, and determining a statistical characteristic value of a candidate segment; and determining whether the current candidate position and the candidate length are correct or not based on the statistical feature values of the candidate segments and a first preset threshold value, and if so, decoding the watermark bit value of each candidate segment, and determining watermark information. And the candidate position and the candidate length of the audio information embedded with the watermark can be quickly and accurately determined, so that the watermark information can be accurately extracted.

Description

Technical Field

[0001] This application relates to the technical field of digital audio watermark processing, and in particular, to a digital audio watermark method, device, electronic device, and storage medium. Background Technique

[0002] Digital audio watermark is a type of digital watermark, mainly used to protect the copyright of audio works. Its principle is to utilize the redundancy of audio signals and the masking effect of the human ear to embed watermark information into the host audio without affecting the perceived quality of the audio signal, so as to achieve purposes such as copyright protection, reliability, and integrity identification of the audio. Like other digital watermark technologies, one problem that must be considered in the actual application of audio watermark technology is various attacks that the watermark may be subjected to. For digital audio watermarks, synchronization attack is a powerful attack method. This attack disrupts the normal timing relationship of the audio, resulting in the inability to locate and retrieve the corresponding embedding interval when the watermark is embedded, thus causing the watermark extraction to fail. Current anti-synchronization audio watermarking schemes often rely on synchronization codes to achieve. One is that the effective payload is reduced because a part of the embedding capacity needs to be set aside for the synchronization code. The other is that the extraction process is too complex, and the sliding window matching of the synchronization code needs to be considered outside the double loop. Therefore, how to perform digital audio watermarking has become a technical problem that cannot be underestimated. Summary of the Invention

[0003] In view of this, the purpose of this application is to provide a digital audio watermark method, device, electronic device, and storage medium. The digital audio watermark method provided by this application has a built-in verification function, can eliminate the synchronization code in the anti-synchronization algorithm, and can quickly and accurately determine the correct candidate positions and candidate lengths of the audio information after the watermark is embedded, so as to accurately extract the watermark information.

[0004] The embodiment of this application provides a digital audio watermark method, and the digital audio watermark method includes: Perform segmentation processing on the first audio information to generate multiple audio segments, perform transform domain processing on each audio segment, and determine the watermark data embedding domain in each audio segment; Based on the digital audio watermark embedding algorithm, perform digital watermark information embedding processing on the audio data in each watermark data embedding domain, determine multiple audio segments after the watermark is embedded, and combine the multiple audio segments after the watermark is embedded to determine the second audio information; Based on a preset plurality of candidate positions and candidate segment lengths, perform segmentation processing and eigenvalue statistical processing on the second audio information, and determine the statistical eigenvalue of each candidate segment; Based on the statistical feature values of each candidate segment, the first preset threshold, and the second preset threshold, determine whether the current candidate position and candidate length are correct. If so, decode the watermark bit values of each candidate segment in sequence based on the correct candidate position and correct candidate length to determine the watermark information of the second audio information.

[0005] In a possible implementation manner, the digital watermark information embedding process is performed on the audio data in each watermark data embedding domain based on the digital audio watermark embedding algorithm to determine multiple audio segments embedded with watermarks, including: If the watermark bit value is 1, add a constant value to the audio data of the first sub-audio segment in two adjacent sub-audio segments corresponding to the watermark data embedding domain, and subtract the constant value from the audio data in the second sub-audio segment to complete the digital watermark information embedding process; If the watermark bit value is 0, add a constant value to the audio data of the first sub-audio segment in two adjacent sub-audio segments corresponding to the watermark data embedding domain, and subtract the constant value from the audio data in the second sub-audio segment to complete the digital watermark information embedding process.

[0006] In a possible implementation manner, for each candidate position and candidate segment length, the second audio information is segmented and the feature value is statistically processed based on a preset plurality of candidate positions and candidate segment lengths to determine the statistical feature value of each candidate segment, including: Segment the second audio information based on the candidate position and candidate segment length to determine multiple candidate segments of the second audio information; For each candidate segment, divide the audio data in the candidate segment into equal first audio data set and second audio data set, calculate the average value of the first audio data set and the second audio data set respectively to determine the first statistical feature value of the first audio data set and the second statistical feature value of the second audio data set.

[0007] In a possible implementation manner, the determination of whether the current candidate position and candidate length are correct based on the statistical feature values of each candidate segment, the first preset threshold, and the second preset threshold includes: Based on the statistical feature value of the candidate segment and the first preset threshold, determine the watermark state of the current candidate segment; wherein, the watermark state includes an invalid watermark state and a valid watermark state, the watermark bit value of the valid watermark state is 0 or 1, and the watermark bit value of the invalid watermark state is 2; Detect whether the number of the invalid watermark states of each candidate segment statistically obtained exceeds the second preset threshold; If so, the current candidate position and candidate length are incorrect; if not, the current candidate position and candidate length are correct.

[0008] In a possible implementation manner, determining the watermark status of the current candidate segment based on the statistical feature value of the candidate segment and the first preset threshold includes: Detecting whether the absolute value of the difference between the first statistical feature value and the second statistical feature value is greater than the first preset threshold; If so, the watermark status of the current candidate segment is a valid watermark status; wherein, if the first statistical feature value is greater than the second statistical feature value, the decoded watermark bit value is 0, and if the first statistical feature value is less than the second statistical feature value, the decoded watermark bit is 1; If not, the watermark status of the current candidate segment is an invalid watermark status.

[0009] In a possible implementation manner, after segmenting the first audio information to generate multiple audio segments, performing transform domain processing on each audio segment, and determining the watermark data embedding domain in each audio segment, the digital audio watermark method further includes: Performing a summation process on the audio data in the watermark data embedding domain to determine a quantized statistical feature value; Determining a quantization step size based on the statistical characteristics of the first audio information and the auditory threshold, and multiplying the quantization step size by a preset fraction to determine a target value; If the watermark bit value is 0, performing a subtraction process on the quantized statistical feature value and the target value; If the watermark bit value is 1, performing an addition process on the quantized statistical feature value and the target value to complete the digital watermark information embedding process.

[0010] In a possible implementation manner, the digital audio watermark method further includes: Segmenting the second audio information based on a preset plurality of candidate positions and candidate segment lengths to determine a plurality of candidate segments, and statistically calculating the statistical feature value of each candidate segment; Performing quantization processing on the statistical feature value of each candidate segment based on the quantization step size to determine the quantization residual value of each candidate segment; If the interval where the quantization residual value is located is the first preset interval or the second preset interval, the watermark status of the candidate segment is a valid watermark status; If the interval where the quantization residual value is located is the third preset interval, the watermark status of the candidate segment is an invalid watermark status; wherein, the ranges of the first preset interval, the third preset interval, and the second preset interval increase in sequence; Based on the number of invalid watermark states of each candidate segment counted and a preset second threshold, determine whether the current candidate position and candidate length are correct.

[0011] An embodiment of the present application further provides a digital audio watermarking device, and the digital audio watermarking device includes: A first processing module, configured to perform segmentation processing on first audio information to generate a plurality of audio segments, perform transform domain processing on each of the audio segments, and determine a watermark data embedding domain in each of the audio segments; A first watermark information embedding module, configured to perform digital watermark information embedding processing on the audio data in each of the watermark data embedding domains based on a digital audio watermark embedding algorithm, determine a plurality of audio segments with embedded watermarks, and combine the plurality of audio segments with embedded watermarks to determine second audio information; A second processing module, configured to perform segmentation processing and eigenvalue statistics processing on the second audio information based on a plurality of preset candidate positions and candidate segment lengths, and determine a statistical eigenvalue of each candidate segment; A first watermark information extraction module, configured to determine whether the current candidate position and candidate length are correct based on the statistical eigenvalue of each candidate segment, a first preset threshold, and a second preset threshold. If so, decode the watermark bit values of each candidate segment in sequence based on the correct candidate position and correct candidate length, and determine the watermark information of the second audio information.

[0012] An embodiment of the present application further provides an electronic device, including: a processor, a memory, and a bus. The memory stores machine-readable instructions executable by the processor. When the electronic device runs, the processor communicates with the memory through the bus. When the machine-readable instructions are executed by the processor, the steps of the digital audio watermarking method as described above are executed.

[0013] An embodiment of the present application further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is run by a processor, the steps of the digital audio watermarking method as described above are executed.

[0014] A digital audio watermarking method, apparatus, electronic device, and storage medium provided by an embodiment of the present application. The digital audio watermarking method includes: performing segmentation processing on first audio information to generate a plurality of audio segments, performing transform domain processing on each of the audio segments to determine a watermark data embedding domain in each of the audio segments; performing digital watermark information embedding processing on the audio data in each of the watermark data embedding domains based on a digital audio watermark embedding algorithm to determine a plurality of audio segments after embedding watermarks, and combining the plurality of audio segments after embedding watermarks to determine second audio information; performing segmentation processing and eigenvalue statistics processing on the second audio information based on a plurality of preset candidate positions and candidate segment lengths to determine a statistical eigenvalue of each candidate segment; determining whether the current candidate position and candidate length are correct based on the statistical eigenvalue of each candidate segment, a first preset threshold, and a second preset threshold. If so, decoding the watermark bit values of each candidate segment in sequence based on the correct candidate position and correct candidate length to determine the watermark information of the second audio information. The digital audio watermarking method provided by the present application has a built-in verification function, can eliminate the synchronization code in the anti-synchronization algorithm, and can quickly and accurately determine the correct candidate position and candidate length of the audio information after embedding the watermark, so as to accurately extract the watermark information.

[0015] To make the above objects, features, and advantages of the present application more obvious and understandable, the following specifically gives preferred embodiments and, in conjunction with the accompanying drawings, makes a detailed description as follows. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] To more clearly illustrate the technical solutions of the embodiments of the present application, the following briefly introduces the accompanying drawings required for the embodiments. It should be understood that the following drawings only show some embodiments of the present application and should not be regarded as limiting the scope. For those of ordinary skill in the art, other related drawings can be obtained based on these drawings without creative efforts.

[0017] Figure 1 It is a flowchart of a digital audio watermarking method provided by an embodiment of the present application; Figure 2 It is a schematic structural diagram of a digital audio watermarking apparatus provided by an embodiment of the present application; Figure 3 It is a schematic structural diagram of a digital audio watermarking apparatus provided by an embodiment of the present application; Figure 4 It is a schematic structural diagram of an electronic device provided by an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0018] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of this application. Apparently, the described embodiments are only a part rather than all of the embodiments of this application. Components of the embodiments of this application described and illustrated herein generally may be arranged and designed in a variety of different configurations. Therefore, the following detailed description of the embodiments of this application provided in the drawings is not intended to limit the scope of the claimed application, but merely represents selected embodiments of this application. Based on the embodiments of this application, every other embodiment obtained by those skilled in the art without creative efforts shall fall within the protection scope of this application.

[0019] First, the applicable application scenarios of this application will be introduced. This application can be applied to the technical field of digital audio watermark processing.

[0020] Through research, it is found that digital audio watermark is a type of digital watermark, mainly used to protect the copyright of audio works. Its principle is to utilize the redundancy of audio signals and the masking effect of the human ear to embed watermark information into the host audio without affecting the perceptual quality of the audio signal, so as to achieve the purposes of copyright protection, reliability, and integrity verification of the audio. Like other digital watermark technologies, one problem that must be considered in the actual application of audio watermark technology is various attacks that the watermark may be subjected to. For digital audio watermarks, synchronization attack is a powerful attack method. This attack disrupts the normal timing relationship of the audio, resulting in the inability to locate and retrieve the corresponding embedding interval when the watermark is embedded, thus causing the failure of watermark extraction. Current anti-synchronization audio watermarking schemes often rely on synchronization codes to achieve, which results in a reduction in the effective payload because a part of the embedding capacity needs to be set aside for the synchronization code. Second, the extraction process is too complex, and the sliding window matching of the synchronization code needs to be considered outside the double loop. Therefore, how to perform digital audio watermarking has become a technical problem that cannot be underestimated.

[0021] Based on this, the embodiments of this application provide a digital audio watermarking method. The digital audio watermarking method provided by this application has a built-in verification function, can eliminate the synchronization code in the anti-synchronization algorithm, and can quickly and accurately determine the correct candidate positions and candidate lengths of the audio information after the watermark is embedded, so as to accurately extract the watermark information.

[0022] Refer to Figure 1 , Figure 1 which is a flowchart of a digital audio watermarking method provided by the embodiments of this application. As shown in Figure 1 , the digital audio watermarking method provided by the embodiments of this application includes: S101: Segment the first audio information to generate multiple audio segments, perform transform domain processing on each of the audio segments, and determine the watermark data embedding domain in each of the audio segments.

[0023] In this step, segment the first audio information to generate multiple audio segments, perform transform domain processing on each audio segment, and determine the watermark data embedding domain in each audio segment.

[0024] Among them, transform domain processing is to convert the audio signal from the time domain (i.e., the original sampling points on the time axis) to other mathematical domains (such as the frequency domain, wavelet domain, etc.) to more efficiently analyze or modify certain characteristics of the signal. Common transform domain processing methods include discrete Fourier transform, discrete cosine transform, discrete wavelet transform, and short-time Fourier transform, etc.

[0025] Here, the first audio information is the original audio information.

[0026] Among them, the purpose of segmenting the first audio information is to embed a watermark bit payload in each audio segment.

[0027] S102: Based on the digital audio watermark embedding algorithm, perform digital watermark information embedding processing on the audio data in each of the watermark data embedding domains, determine multiple audio segments with embedded watermarks, and combine the multiple audio segments with embedded watermarks to determine the second audio information.

[0028] In this step, according to the digital audio watermark embedding algorithm, perform digital watermark information embedding processing on the audio data in each watermark data embedding domain, determine multiple audio segments with embedded watermarks, and combine the multiple audio segments with embedded watermarks to determine the second audio information.

[0029] Here, according to the digital audio watermark embedding algorithm, modify the audio data in each watermark data embedding domain, and make different modifications to the audio data according to whether the watermark load bit is 0 or 1, so that the data carries watermark information.

[0030] Among them, the digital audio watermark embedding algorithm can be a patchwork - type algorithm.

[0031] In a possible implementation manner, the performing digital watermark information embedding processing on the audio data in each of the watermark data embedding domains based on the digital audio watermark embedding algorithm to determine multiple audio segments with embedded watermarks includes: If the watermark bit value is 1, a constant value is added to the audio data of the first sub-audio segment among two adjacent sub-audio segments corresponding to the watermark data embedding domain, and the constant value is subtracted from the audio data in the second sub-audio segment, so as to complete the digital watermark information embedding process; if the watermark bit value is 0, a constant value is added to the audio data of the first sub-audio segment among two adjacent sub-audio segments corresponding to the watermark data embedding domain, and the constant value is subtracted from the audio data in the second sub-audio segment, so as to complete the digital watermark information embedding process.

[0032] Among them, the constant value added and the constant value subtracted should be the same constant.

[0033] Here, the two adjacent sub-audio segments corresponding to the watermark data embedding domain are obtained by dividing the audio segmentation into two equal parts.

[0034] S103: Perform segmentation processing and eigenvalue statistics processing on the second audio information based on a plurality of preset candidate positions and candidate segment lengths, and determine the statistical eigenvalue of each candidate segment.

[0035] In this step, segmentation processing and eigenvalue statistics processing are performed on the second audio information according to a plurality of preset candidate positions and candidate segment lengths, and the statistical eigenvalue of each candidate segment is determined.

[0036] In a possible implementation manner, for each candidate position and candidate segment length, the performing segmentation processing and eigenvalue statistics processing on the second audio information based on a plurality of preset candidate positions and candidate segment lengths, and determining the statistical eigenvalue of each candidate segment includes: A: Perform segmentation processing on the second audio information based on the candidate position and the candidate segment length, and determine a plurality of candidate segments of the second audio information.

[0037] Here, segmentation processing is performed on the second audio information according to the candidate position and the candidate segment length, and a plurality of candidate segments of the second audio information are determined.

[0038] Among them, the role of the candidate position is to use multiple positions as the starting points of the candidate segments.

[0039] B: For each of the candidate segments, divide the audio data in the candidate segment into equal first audio data set and second audio data set, calculate the mean values of the first audio data set and the second audio data set respectively, and determine the first statistical eigenvalue of the first audio data set and the second statistical eigenvalue of the second audio data set.

[0040] Here, for each candidate segment, an equal - sized first audio data set and a second audio data set are divided according to the audio data in the candidate segment, and the mean values of the first audio data set and the second audio data set are calculated respectively to determine the first statistical feature value of the first audio data set and the second statistical feature value of the second audio data set.

[0041] Here, the statistical feature value can also be determined according to energy, variance, singular value, etc., and this part is not specifically limited.

[0042] S104: Determine whether the current candidate position and candidate length are correct based on the statistical feature value of each candidate segment, the first preset threshold, and the second preset threshold. If so, decode the watermark bit values of each candidate segment in sequence based on the correct candidate position and correct candidate length to determine the watermark information of the second audio information.

[0043] In this step, determine whether the current candidate position and candidate length are correct according to the statistical feature value of each candidate segment and the first preset threshold. If so, decode the watermark bit values of each candidate segment in sequence based on the correct candidate position and correct candidate length to determine the watermark information of the second audio information.

[0044] Here, even if the audio segment has been embedded with a watermark, if the segment position or segment length used during decoding is incorrect, the decoding process will still regard the signal as random data. The reasons are as follows: If segmentation starts from the wrong position, the rules for embedding the watermark (such as the mean difference of the Patchwork - like algorithm or the quantization value of the QIM - like algorithm) will be violated, resulting in the decoding result not being able to match the valid watermark state (0 or 1). If the wrong segment length is used, the data with the embedded watermark will be divided into wrong small segments, thus violating the embedding rules. Therefore, it is necessary to determine the correct candidate position and candidate length.

[0045] In a possible implementation manner, the determining whether the current candidate position and candidate length are correct based on the statistical feature value of each candidate segment, the first preset threshold, and the second preset threshold includes: a: Determine the watermark state of the current candidate segment based on the statistical feature value of the candidate segment and the first preset threshold; wherein, the watermark state includes an invalid watermark state and a valid watermark state, the watermark bit value of the valid watermark state is 0 or 1, and the watermark bit value of the invalid watermark state is 2.

[0046] Here, determine the watermark state of the current candidate segment according to the statistical feature value of the candidate segment and the first preset threshold.

[0047] Among them, the innovative designed watermark states in this application include an invalid watermark state and a valid watermark state. The watermark bit values of all valid watermark states are 0 or 1, and the watermark bit value of the invalid watermark state is 2.

[0048] In a possible implementation manner, determining the watermark state of the current candidate segment based on the statistical feature value of the candidate segment and the first preset threshold includes: (1): Detect whether the absolute value of the difference between the first statistical feature value and the second statistical feature value is greater than the first preset threshold.

[0049] Here, detect whether the absolute value of the difference between the first statistical feature value and the second statistical feature value is greater than the first preset threshold.

[0050] Among them, the first preset threshold is determined according to expert experience.

[0051] (2): If so, the watermark state of the current candidate segment is a valid watermark state; among them, if the first statistical feature value is greater than the second statistical feature value, the decoded watermark bit value is 0, and if the first statistical feature value is less than the second statistical feature value, the decoded watermark bit is 1; if not, the watermark state of the current candidate segment is an invalid watermark state.

[0052] Here, if so, the watermark state of the current candidate segment is a valid watermark state; if not, the watermark state of the current candidate segment is an invalid watermark state.

[0053] Among them, the three-state watermark algorithm proposed in this application transforms the patchwork method into 3 categories, as follows:

[0054] Among them, when the difference between the two calculated values and is large (greater than the first preset threshold ), the watermark is valid. According to the and relative size relationship, the corresponding watermark bit is 0 or 1; when the absolute value of the difference between and is less than the threshold , it means that the watermark is invalid, that is, there is no watermark embedded in the current data segment.

[0055] b: Detect whether the number of the invalid watermark states of each candidate segment counted exceeds the second preset threshold; if so, the current candidate position and candidate length are incorrect, and if not, the current candidate position and candidate length are correct.

[0056] Here, it is detected whether the number of invalid watermark states of each candidate segment detected and statistically counted exceeds a second preset threshold; if so, the current candidate position and candidate length are incorrect, and the candidate position and candidate length are continuously determined; if not, the current candidate position and candidate length are correct.

[0057] Among them, the second preset threshold is determined according to expert experience.

[0058] In this application, the new three-state watermark cannot simplify the dual search, and multiple positions still need to be considered with multiple segment lengths for each position. However, the new algorithm can simplify the subsequent synchronization code matching. In fact, since the new algorithm has a built-in verification function, the synchronization code can be completely removed, and all the embedded data are valid payloads without setting aside a part of the embedding capacity for the synchronization code. For the new algorithm, only at each candidate position, multiple segments are continuously decoded using the candidate segment lengths. Then, it is only necessary to count how many '2's appear in the decoding results. According to the previous algorithm description, for an audio segment without a watermark or an audio segment with a watermark but incorrect segment parameters, the proportion of '2' after decoding is relatively large. If the number of watermark bit values '2' after decoding is less than or equal to the second preset threshold, it indicates that the correct segment position and segment length are found; otherwise, it indicates that the parameters of the segment length or segment position are incorrect.

[0059] In a possible implementation manner, after segmenting the first audio information to generate multiple audio segments, performing transform domain processing on each of the audio segments, and determining the watermark data embedding domain in each of the audio segments, the digital audio watermark method further includes: I: Performing a summation process on the audio data in the watermark data embedding domain to determine a quantized statistical eigenvalue.

[0060] Here, a summation process is performed on the audio data in the watermark data embedding domain to determine a quantized statistical eigenvalue.

[0061] II: Determining a quantization step based on the statistical characteristics and auditory threshold of the first audio information, and multiplying the quantization step by a preset fraction to determine a target value.

[0062] Here, a quantization step Δ is determined according to the statistical characteristics and auditory threshold of the first audio information, and the quantization step is multiplied by a preset fraction to determine a target value Δ / 6.

[0063] III: If the watermark bit value is 0, subtracting the target value from the quantized statistical eigenvalue; if the watermark bit value is 1, adding the target value to the quantized statistical eigenvalue to complete the digital watermark information embedding process.

[0064] Here, if the watermark bit value is 0, subtract the quantized statistical eigenvalue from the target value; if the watermark bit value is 1, add the quantized statistical eigenvalue to the target value, so as to complete the digital watermark information embedding process.

[0065] In this application, a three-state watermark algorithm proposed based on the original QIM algorithm transforms the above method to output three categories. When embedding the digital watermark, after quantization, according to whether the watermark bit is 0 or 1, add -Δ / 6 or Δ / 6 to the quantization result. In a possible implementation manner, the digital audio watermark method further includes: i: Segment the second audio information based on a preset plurality of candidate positions and candidate segment lengths to determine a plurality of candidate segments, and calculate the statistical eigenvalue of each candidate segment.

[0066] Here, segment the second audio information based on a preset plurality of candidate positions and candidate segment lengths to determine a plurality of candidate segments, and calculate the statistical eigenvalue of each candidate segment.

[0067] ii: Quantize the statistical eigenvalue of each candidate segment based on the quantization step size to determine the quantization residual value of each candidate segment.

[0068] Here, quantize the statistical eigenvalue of each candidate segment based on the quantization step size to determine the quantization residual value of each candidate segment.

[0069] iii: If the interval where the quantization residual value is located is the first preset interval or the second preset interval, the watermark state of the candidate segment is a valid watermark state; if the interval where the quantization residual value is located is the third preset interval, the watermark state of the candidate segment is an invalid watermark state; wherein, the ranges of the first preset interval, the third preset interval, and the second preset interval increase in sequence.

[0070] Here, if the interval where the quantization residual value is located is the first preset interval or the second preset interval, the watermark state of the candidate segment is a valid watermark state; if the interval where the quantization residual value is located is the third preset interval, the watermark state of the candidate segment is an invalid watermark state.

[0071] Among them, the first preset interval is (0, 2 / 6 ], the second preset interval is (2 / 6, 4 / 6], and the third preset interval is (4 / 6, 1].

[0072] iv: Based on the number of invalid watermark states of each candidate segment counted and the second preset threshold, determine whether the current candidate position and candidate length are correct.

[0073] Here, according to the number of invalid watermark states of each candidate segment counted and a preset second threshold, it is determined whether the current candidate position and candidate length are correct.

[0074] A digital audio watermarking method provided by an embodiment of the present application, the digital audio watermarking method includes: performing segmentation processing on first audio information to generate a plurality of audio segments, performing transform domain processing on each of the audio segments to determine a watermark data embedding domain in each of the audio segments; performing digital watermark information embedding processing on the audio data in each of the watermark data embedding domains based on a digital audio watermark embedding algorithm to determine a plurality of audio segments after embedding watermarks, and combining the plurality of audio segments after embedding watermarks to determine second audio information; performing segmentation processing and eigenvalue statistics processing on the second audio information based on a plurality of preset candidate positions and candidate segment lengths to determine a statistical eigenvalue of each candidate segment; determining whether the current candidate position and candidate length are correct based on the statistical eigenvalue of each candidate segment, a first preset threshold, and a second preset threshold. If so, decoding the watermark bit values of each candidate segment in sequence based on the correct candidate position and correct candidate length to determine the watermark information of the second audio information. The digital audio watermarking method provided by the present application has a built-in verification function, can eliminate the synchronization code in the anti-synchronization algorithm, and can quickly and accurately determine the correct candidate position and candidate length of the audio information after embedding the watermark, so as to accurately extract the watermark information.

[0075] Please refer to Figure 2 、 Figure 3 , Figure 2 is one of the structural schematic diagrams of a digital audio watermarking device provided by an embodiment of the present application; Figure 3 is the second of the structural schematic diagrams of a digital audio watermarking device provided by an embodiment of the present application. As Figure 2 shown in A first processing module 210, configured to perform segmentation processing on first audio information to generate a plurality of audio segments, perform transform domain processing on each of the audio segments, and determine a watermark data embedding domain in each of the audio segments; A first watermark information embedding module 220, configured to perform digital watermark information embedding processing on the audio data in each of the watermark data embedding domains based on a digital audio watermark embedding algorithm, determine a plurality of audio segments after embedding watermarks, and combine the plurality of audio segments after embedding watermarks to determine second audio information; A second processing module 230, configured to perform segmentation processing and eigenvalue statistics processing on the second audio information based on a plurality of preset candidate positions and candidate segment lengths, and determine a statistical eigenvalue of each candidate segment; The first watermark information extraction module 240 is configured to determine whether the current candidate position and candidate length are correct based on the statistical feature values of each candidate segment, the first preset threshold, and the second preset threshold. If so, the watermark bit values of each candidate segment are decoded in sequence based on the correct candidate position and the correct candidate length, and the watermark information of the second audio information is determined.

[0076] Further, when the first watermark information embedding module 220 is used to perform digital watermark information embedding processing on the audio data in each watermark data embedding domain based on the digital audio watermark embedding algorithm and determine a plurality of audio segments with embedded watermarks, the first watermark information embedding module 220 is specifically configured to: If the watermark bit value is 1, a constant value is added to the audio data of the first sub-audio segment in two adjacent sub-audio segments corresponding to the watermark data embedding domain, and a constant value is subtracted from the audio data in the second sub-audio segment to complete the digital watermark information embedding processing; If the watermark bit value is 0, a constant value is added to the audio data of the first sub-audio segment in two adjacent sub-audio segments corresponding to the watermark data embedding domain, and a constant value is subtracted from the audio data in the second sub-audio segment to complete the digital watermark information embedding processing.

[0077] Further, when the second processing module 230 is used to perform segmentation processing and eigenvalue statistics processing on the second audio information based on a plurality of preset candidate positions and candidate segment lengths for each candidate position and candidate segment length, and determine the statistical feature value of each candidate segment, the second processing module 230 is specifically configured to: Perform segmentation processing on the second audio information based on the candidate position and the candidate segment length to determine a plurality of candidate segments of the second audio information; For each candidate segment, divide the first audio data set and the second audio data set of equal parts based on the audio data in the candidate segment, calculate the mean values of the first audio data set and the second audio data set respectively, and determine the first statistical feature value of the first audio data set and the second statistical feature value of the second audio data set.

[0078] Further, when the first watermark information extraction module 240 is used to determine whether the current candidate position and candidate length are correct based on the statistical feature values of each candidate segment, the first preset threshold, and the second preset threshold, the first watermark information extraction module 240 is specifically configured to: Based on the statistical eigenvalue of the candidate segment and the first preset threshold, determine the watermark status of the current candidate segment; wherein, the watermark status includes an invalid watermark status and a valid watermark status, the watermark bit value of the valid watermark status is 0 or 1, and the watermark bit value of the invalid watermark status is 2; Detect whether the number of the invalid watermark status of each candidate segment statistically obtained exceeds the second preset threshold; If so, the current candidate position and candidate length are incorrect; if not, the current candidate position and candidate length are correct.

[0079] Further, when the first watermark information extraction module 240 is used to determine the watermark status of the current candidate segment based on the statistical eigenvalue of the candidate segment and the first preset threshold, the first watermark information extraction module 240 is specifically used for: Detect whether the absolute value of the difference between the first statistical eigenvalue and the second statistical eigenvalue is greater than the first preset threshold; If so, the watermark status of the current candidate segment is a valid watermark status; wherein, if the first statistical eigenvalue is greater than the second statistical eigenvalue, the decoded watermark bit value is 0, and if the first statistical eigenvalue is less than the second statistical eigenvalue, the decoded watermark bit is 1; If not, the watermark status of the current candidate segment is an invalid watermark status.

[0080] Further, as Figure 3 shown, the digital audio watermark device 300 further includes a second watermark information embedding module 250, and the second watermark information embedding module 250 is used for: Perform a summation process on the audio data in the watermark data embedding domain to determine a quantized statistical eigenvalue; Based on the statistical characteristics of the first audio information and the auditory threshold, determine a quantization step size, multiply the quantization step size by a preset fraction to determine a target value; If the watermark bit value is 0, perform a subtraction process on the quantized statistical eigenvalue and the target value; If the watermark bit value is 1, perform an addition process on the quantized statistical eigenvalue and the target value to complete the digital watermark information embedding process.

[0081] Further, as Figure 3 shown, the digital audio watermark device 300 further includes a second watermark information extraction module 260, and the second watermark information extraction module 260 is used for: Based on a plurality of preset candidate positions and candidate segment lengths, perform a segmentation process on the second audio information to determine a plurality of candidate segments, and statistically obtain the statistical eigenvalue of each candidate segment; Quantize the statistical feature values of each candidate segment based on the quantization step size to determine the quantization residual value of each candidate segment; If the interval where the quantization residual value is located is the first preset interval or the second preset interval, the watermark state of the candidate segment is a valid watermark state; If the interval where the quantization residual value is located is the third preset interval, the watermark state of the candidate segment is an invalid watermark state; wherein, the ranges of the first preset interval, the third preset interval, and the second preset interval increase in sequence; Based on the number of invalid watermark states of each candidate segment and a preset second threshold, determine whether the current candidate position and candidate length are correct.

[0082] A digital audio watermarking device provided by an embodiment of the present application, the digital audio watermarking device includes: a first processing module, configured to segment first audio information to generate multiple audio segments, perform transform domain processing on each of the audio segments to determine a watermark data embedding domain in each of the audio segments; a first watermark information embedding module, configured to perform digital watermark information embedding processing on the audio data in each of the watermark data embedding domains based on a digital audio watermark embedding algorithm to determine multiple audio segments with embedded watermarks, and combine the multiple audio segments with embedded watermarks to determine second audio information; a second processing module, configured to perform segmentation processing and eigenvalue statistics processing on the second audio information based on a preset multiple of candidate positions and candidate segment lengths to determine the statistical feature values of each candidate segment; a first watermark information extraction module, configured to determine whether the current candidate position and candidate length are correct based on the statistical feature values of each candidate segment, a first preset threshold, and a second preset threshold. If so, decode the watermark bit values of each candidate segment in sequence based on the correct candidate position and the correct candidate length to determine the watermark information of the second audio information. The digital audio watermarking method provided by the present application has a built-in verification function, can eliminate the synchronization code in the anti-synchronization algorithm, and can quickly and accurately determine the correct candidate position and candidate length of the audio information after embedding the watermark, so as to accurately extract the watermark information.

[0083] Please refer to Figure 4 , Figure 4 which is a schematic structural diagram of an electronic device provided by an embodiment of the present application. As Figure 4 shown in

[0084] The memory 420 stores machine-readable instructions executable by the processor 410. When the electronic device 400 runs, the processor 410 communicates with the memory 420 via a bus 430. When the machine-readable instructions are executed by the processor 410, the steps of the digital audio watermarking method in the method embodiments as described above can be executed. For the specific implementation manners, reference can be made to the method embodiments and will not be elaborated herein. Figure 1 shown in the method embodiments, and for the specific implementation manners, reference can be made to the method embodiments and will not be elaborated herein.

[0085] The embodiments of the present application further provide a computer-readable storage medium. A computer program is stored on the computer-readable storage medium. When the computer program is run by a processor, the steps of the digital audio watermarking method in the method embodiments as described above can be executed. For the specific implementation manners, reference can be made to the method embodiments and will not be elaborated herein. Figure 1 shown in the method embodiments, and for the specific implementation manners, reference can be made to the method embodiments and will not be elaborated herein.

[0086] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0087] In several embodiments provided by the present application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division, and there can be other division manners in actual implementation. For another example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some communication interfaces. The indirect couplings or communication connections of the devices or units can be in electrical, mechanical, or other forms.

[0088] The units described as separate components may or may not be physically separated. The components displayed as units may or may not be physical units, that is, they can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0089] In addition, in each embodiment of the present application, the functional units can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit.

[0090] When the above-mentioned functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a non-volatile computer-readable storage medium executable by a processor. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of this application. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM), random access memories (RAM), magnetic disks, or optical discs that can store program codes.

[0091] Finally, it should be noted that the above-mentioned embodiments are only specific implementation manners of this application, used to illustrate the technical solution of this application, rather than limiting it. The protection scope of this application is not limited thereto. Although this application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: any person skilled in the art within the technical scope disclosed in this application can still modify the technical solutions recorded in the foregoing embodiments or can easily think of changes, or perform equivalent replacements on some of the technical features; and these modifications, changes, or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of this application and should all be covered by the protection scope of this application. Therefore, the protection scope of this application should be subject to the protection scope of the claims.

Claims

1. A digital audio watermarking method, characterized in that: The digital audio watermark method comprises: Segmenting the first audio information to generate a plurality of audio segments, performing transform domain processing on each of the audio segments, and determining a watermark data embedding domain in each of the audio segments; Performing digital watermark information embedding processing on the audio data in each watermark data embedding domain based on a digital audio watermark embedding algorithm, determining a plurality of watermarked audio segments, and combining the plurality of watermarked audio segments to determine second audio information; Based on a plurality of preset candidate positions and candidate segment lengths, the second audio information is segmented and feature value statistically processed to determine a statistical feature value of each candidate segment; Based on the statistical characteristic value of each candidate segment, the first preset threshold and the second preset threshold, determine whether the current candidate position and candidate length are correct. If so, decode the watermark bit value of each candidate segment in turn based on the correct candidate position and the correct candidate length to determine the watermark information of the second audio information.

2. The digital audio watermarking method according to claim 1, characterized in that: The method of performing digital watermark information embedding processing on each audio data in the watermark data embedding domain based on a digital audio watermark embedding algorithm to determine a plurality of audio segments after watermark embedding includes: If the watermark bit value is 1, a constant value is added to the audio data of the first sub-audio segment of the two adjacent sub-audio segments corresponding to the watermark data embedding domain, and the constant value is subtracted from the audio data of the second sub-audio segment, so as to complete the digital watermark information embedding process; If the watermark bit value is 0, a constant value is added to the audio data of the first sub-audio segment of the two adjacent sub-audio segments corresponding to the watermark data embedding domain, and the constant value is subtracted from the audio data in the second sub-audio segment to complete the digital watermark information embedding process.

3. The digital audio watermarking method according to claim 1, characterized in that: For each candidate position and candidate segment length, the second audio information is segmented and feature value statistically processed based on the preset multiple candidate positions and candidate segment lengths to determine the statistical feature value of each candidate segment, including: Segment the second audio information based on the candidate positions and the candidate segment lengths to determine a plurality of candidate segments of the second audio information; For each of the candidate segments, a first audio data set and a second audio data set are divided into equal parts based on the audio data in the candidate segment, and the means are calculated for the first audio data set and the second audio data set, respectively, to determine a first statistical feature value of the first audio data set and a second statistical feature value of the second audio data set.

4. The digital audio watermarking method according to claim 3, characterized in that: The determining whether the current candidate position and the candidate length are correct based on the statistical feature value of each candidate segment, the first preset threshold and the second preset threshold includes: Based on the statistical feature value of the candidate segment and the first preset threshold, determine the watermark state of the current candidate segment; wherein the watermark state includes an invalid watermark state and a valid watermark state, the watermark bit value of the valid watermark state is 0 or 1, and the watermark bit value of the invalid watermark state is 2; Detecting whether the number of the invalid watermark states of each candidate segment counted exceeds a second preset threshold; If so, the current candidate position and candidate length are incorrect; if not, the current candidate position and candidate length are correct.

5. The digital audio watermarking method according to claim 4, characterized in that: The determining of the watermark state of the current candidate segment based on the statistical feature value of the candidate segment and the first preset threshold comprises: Detecting whether an absolute value of a difference between the first statistical characteristic value and the second statistical characteristic value is greater than a first preset threshold; If yes, the watermark state of the current candidate segment is a valid watermark state; wherein, if the first statistical characteristic value is greater than the second statistical characteristic value, the decoded watermark bit value is 0, and if the first statistical characteristic value is less than the second statistical characteristic value, the decoded watermark bit value is 1; If not, the watermark state of the current candidate segment is an invalid watermark state.

6. The digital audio watermarking method according to claim 1, characterized in that: After the first audio information is segmented to generate a plurality of audio segments, each of the audio segments is transformed into a domain, and a watermark data embedding domain in each of the audio segments is determined, the digital audio watermark method further comprises: Performing summation processing on the audio data in the watermark data embedding domain to determine a quantized statistical characteristic value; Determine a quantization step length based on the statistical characteristics of the first audio information and the auditory threshold, and multiply the quantization step length by a preset score to determine a target value; If the watermark bit value is 0, subtracting the quantized statistical characteristic value from the target value; If the watermark bit value is 1, the quantized statistical characteristic value and the target value are added to complete the digital watermark information embedding process.

7. The digital audio watermarking method according to claim 6, characterized in that: The digital audio watermark method further comprises: Segmenting the second audio information based on a plurality of preset candidate positions and candidate segment lengths to determine a plurality of candidate segments, and calculating a statistical feature value of each candidate segment; Quantizing the statistical characteristic value of each candidate segment based on the quantization step size to determine the quantized residual value of each candidate segment; If the interval in which the quantized residual value is located is the first preset interval or the second preset interval, the watermark state of the candidate segment is a valid watermark state; If the interval in which the quantized residual value is located is a third preset interval, the watermark state of the candidate segment is an invalid watermark state; wherein the ranges of the first preset interval, the third preset interval and the second preset interval are successively larger; Based on the counted number of invalid watermark states of each candidate segment and the preset second threshold, it is determined whether the current candidate position and candidate length are correct.

8. A digital audio watermarking device, characterized in that: The digital audio watermark device comprises: A first processing module, configured to perform segment processing on the first audio information to generate a plurality of audio segments, perform transform domain processing on each of the audio segments, and determine a watermark data embedding domain in each of the audio segments; A first watermark information embedding module is used to perform digital watermark information embedding processing on each audio data in the watermark data embedding domain based on a digital audio watermark embedding algorithm, determine a plurality of audio segments after watermark embedding, and combine the plurality of audio segments after watermark embedding to determine the second audio information; A second processing module, configured to perform segmentation processing and feature value statistical processing on the second audio information based on a plurality of preset candidate positions and candidate segment lengths, and determine a statistical feature value of each candidate segment; The first watermark information extraction module is used to determine whether the current candidate position and candidate length are correct based on the statistical feature value of each candidate segment, the first preset threshold and the second preset threshold. If so, the watermark bit value of each candidate segment is decoded in turn based on the correct candidate position and the correct candidate length to determine the watermark information of the second audio information.

9. An electronic device, characterized in that: include: A processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor and the memory communicate through the bus, and the machine-readable instructions are executed by the processor to execute the steps of the digital audio watermarking method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the digital audio watermarking method according to any one of claims 1 to 7 are executed.

Images