JP2000049616A - Signal processor, record medium and signal processing method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To bury desired data such as a watermark in the compression process of information data without reducing an information amount and to prevent illegal copying by simple constitution by burying input data in coefficient data obtained by orthogonal transformation. SOLUTION: At the time of recording digital audio signals DA1 from satellite broadcasting or the like in a mini disk or the like, the prescribed bit of prescribed coefficient data among the plural coefficient data obtained by performing an orthogonal transformation processing by an orthogonal transformation part 12 in a data compression process is set to a logical value corresponding to the data DC of a user watermark and buried in an encoding part 13. Then, after performing an encoding processing and compressing the data by a signal compression part 14, a recording processing is performed. Thus in a data expansion process at the time of reproduction, the data of the user watermark are decoded, the illegal copying is detected, secondary duplication is inhibited and further illegal copying is prevented.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a signal processing device, a recording medium, and a signal processing method. Can be. The present invention embeds input data such as a user watermark in coefficient data generated by orthogonally transforming an input signal, thereby achieving ATRAC (Adaptive Tracing).
nsform Acoustic Coding), MPEG (Moving Picture)
It is an object of the present invention to propose a signal processing device and a signal processing method capable of embedding user watermark data and the like in a process of compressing data by an Experts Group, etc., and a recording medium using the signal processing device and the signal processing method.

[0002]

2. Description of the Related Art Conventionally, there has been proposed a method for protecting a copyright of a recording medium such as an optical disk by a so-called watermark. In this method, for example, a modulation signal or the like of data related to copyright is superimposed on an audio signal or the like at a minute signal level that does not affect reproduction of an audio signal or the like.
It can also be used in digital broadcasting and the like.

[0003]

By the way, in the near future, C
In S digital broadcasting and the like, it is conceivable that various contents are downloaded, recorded on a mini-disc or the like, and enjoyed by individuals. In such a content distribution channel, in addition to a source watermark for embedding copyright information and the like in an original source, a user watermark for embedding user ID information and the like of each individual who uses the source is considered necessary.

[0004] In this case, in CS digital broadcasting, each user can use an IRD (Integrated Receiver Decoder).
), It is necessary to add a user watermark using a device such as an IRD.

In this case, for example, ATRAC, MPEG
If the watermark can be embedded in the process of compressing the data by using the method described above, the overall configuration can be simplified accordingly. At this time, if the watermark can be embedded in each compression process, it is considered that the reduction in the amount of information due to the compression processing can be avoided, and the compression resistance can be improved accordingly.

The present invention has been made in consideration of the above points, and a signal processing apparatus and a signal processing method capable of embedding data such as a user watermark in the course of data compression by ATRAC, MPEG, etc. An apparatus and a recording medium using a signal processing method are to be proposed.

[0007]

According to the present invention, there is provided a signal processing apparatus comprising: an orthogonal transform unit for orthogonally transforming an input signal to generate coefficient data;
The apparatus includes a coefficient data processing unit that embeds input data in predetermined coefficient data of coefficient data to generate correction data of the coefficient data, and an encoding unit that encodes the correction data of the coefficient data.

Decoding means for decoding an input signal to output coefficient data; coefficient data processing means for detecting predetermined data embedded in the coefficient data; Conversion means.

In the recording medium, of the coefficient data generated by performing orthogonal transformation processing on the input signal, the input data is embedded in predetermined coefficient data to generate correction data of the coefficient data. Is encoded and recorded.

Further, in the signal processing method, the input signal is subjected to orthogonal transformation processing to generate coefficient data, and the input data is embedded in predetermined coefficient data of the coefficient data to generate correction data of the coefficient data. Encode the correction data.

In the signal processing method, the input signal is decoded to generate coefficient data, predetermined data embedded in the coefficient data is detected, and the coefficient data is subjected to inverse orthogonal transform processing and output.

In many data compression processes, an orthogonal transform unit for generating coefficient data by performing an orthogonal transform process on an input signal and an encoding process unit for encoding the coefficient data are provided. In this way, by providing coefficient data processing means for generating correction data of the coefficient data by embedding the input data in predetermined coefficient data of the coefficient data, the input data such as the user watermark is embedded in the data compression process. be able to.

[0013] The input signal thus compressed is decoded into an original input signal by decoding means for decoding and outputting coefficient data and inverse orthogonal transform means for subjecting the coefficient data to inverse orthogonal transform processing and outputting the result. Data is decompressed. Therefore, by providing the coefficient data processing means for detecting predetermined data embedded in the coefficient data, data such as the user watermark embedded in the data compression process can be decoded in the data decompression process. .

In the recording medium, of the coefficient data generated by performing an orthogonal transformation process on the input signal, the input data is embedded in predetermined coefficient data to generate correction data of the coefficient data. If the data is encoded and recorded, a recording medium in which copyright data and the like are embedded in the data compression process can be obtained.

Further, in the signal processing method, the input signal is subjected to orthogonal transformation processing to generate coefficient data, and the input data is embedded in predetermined coefficient data of the coefficient data to generate correction data of the coefficient data. If the correction data is encoded, it is possible to embed data such as a user watermark in the data compression process of orthogonally transforming and encoding.

Further, in the signal processing method, if the input signal is decoded to generate coefficient data, predetermined data embedded in the coefficient data is detected, and the coefficient data is subjected to inverse orthogonal transform processing and output, the above-described method is obtained. The data such as the user watermark embedded in the data compression process can be decoded in the data expansion process.

[0017]

Embodiments of the present invention will be described below in detail with reference to the drawings.

(1) First Embodiment (1-1) Configuration of First Embodiment FIG. 2 is a block diagram showing an entire system for detecting unauthorized copy according to a first embodiment of the present invention. It is. In this embodiment, after a CS digital broadcast wave, which is an audio source, is uplinked from a broadcast station 2, a satellite 3
The data is further downlinked and received by the general home 4. At this time, the general household 4 receives the CS digital broadcast wave by the IRD 5, and records a digital audio signal based on the CS digital broadcast wave on a recording medium such as the mini disk 6 by, for example, a magneto-optical disk device. Furthermore, during the data compression process in this magneto-optical disk device,
The data DC of the user watermark according to each user, the IRD, the user ID assigned to the magneto-optical disk or the like is embedded in the digital audio signal.

Thus, it is legal to personally enjoy the mini-disc 6 on which the digital audio signal is recorded, but if the mini-disc 6 is sold in the market, it is subject to unauthorized copying. Thereby, in this system, in the medium 6A on which the digital audio signal is recorded in this way and which is distributed in the market,
In the process of reproducing the digital audio signal, the decoder 7
To detect the data DC of the user watermark, thereby pursuing the source of the illegal copy.

FIG. 1 is a block diagram showing an encoder in the IRD 5. The IRD 5 has a built-in optical disk device and records a digital audio signal by CS digital broadcasting on an optical disk in response to a user operation. At this time, the IRD 5 compresses the data of the digital audio signal DA1 by the encoder 10 and records it on the optical disk, and embeds the data DC of the user watermark into the digital audio signal DA1 at the time of this data compression.

That is, in the encoder 10, the pre-processing section 11 performs pre-processing such as band limiting processing and signal level correction processing corresponding to the encoding processing of the digital audio signal DA1, and
Outputs 1. The orthogonal transformer 12 orthogonally transforms the digital audio signal DA1 output from the preprocessor 11, and outputs coefficient data obtained as a result. The encoding unit 13 sets predetermined bits of the coefficient data of the predetermined frequency in the coefficient data to the logical value of the data DC of the user watermark, thereby embedding the user watermark data in the coefficient data and outputting the data. The signal compression section 14 performs an encoding process on the coefficient data output from the encoding section 13, thereby compressing and outputting the coefficient data.

FIG. 3 is a flowchart showing a processing procedure in the encoder 10. The encoder 10
The process moves from step SP1 to step SP2, and reads the digital audio signal DA1 based on the time-series sample data. Subsequently, in step SP3, the encoder 10 performs preprocessing such as band division and gain correction specific to each compression standard on the digital audio signal DA1.

Subsequently, the encoder 10 proceeds to step SP4
Then, the digital audio signal DA1 is orthogonally transformed by MDCT calculation or the like, and user watermark data is embedded in step SP5 following the coefficient data obtained as a result. Subsequently, the encoder 10
In step SP6, encoding processing by signal compression processing such as vector quantization and adaptive encoding peculiar to each compression standard is executed, and in subsequent step SP7, data by this encoding processing is recorded and output in, for example, a buffer memory or the like. I do.

In the following step SP8, the encoder 10 determines whether or not the data of the digital audio signal DA1 is completed. If a negative result is obtained here, the process returns to step SP2, whereas if a positive result is obtained. ,
The process moves to step SP9, and this processing procedure ends.

FIG. 4 is a block diagram showing the decoder 7. The decoder 7 decompresses the data-compressed digital audio signal DA3 obtained by reproducing the mini disc 6 and the like, and detects the user watermark data DC from the digital audio signal DA3.

That is, in the decoder 7, the signal decompression unit 21 is a signal processing circuit corresponding to the signal compression unit 14 in FIG. 1 and decodes the digital audio signal DA3 to generate original coefficient data. Decoding section 22
Is a signal processing circuit corresponding to the encoding unit 13 shown in FIG. 1. By sequentially extracting predetermined bits of predetermined coefficient data from the coefficient data, data DC of the user watermark assigned to the coefficient data is obtained. Detect and output.

The inverse orthogonal transform unit 23 is an orthogonal transform unit 1 shown in FIG.
2, which performs inverse orthogonal transform processing on the coefficient data and outputs the result. The post-processing unit 24 is a signal processing circuit corresponding to the pre-processing unit 11 of FIG. 1, and outputs a digital audio signal DA4 generated by performing an inverse orthogonal transform process by performing a band combining process, a signal level correcting process, and the like.

FIG. 5 is a flowchart showing the processing procedure of the decoder 7. In the decoder 7, the process proceeds from step SP10 to step SP11, in which the digital audio signal DA3 to be reproduced is input for each sample block, and then decoded in step SP12 to generate coefficient data. Here, this decoding process is executed by spectrum decoding, inverse quantization, etc., according to the format of the data compression.

Subsequently, the decoder 7 proceeds to step SP13, extracts predetermined bits of predetermined coefficient data from the coefficient data generated in this way, and thereby detects user watermark data DC. Subsequently, the process proceeds to step SP14, where the decoder 7 performs an inverse orthogonal transform process on the coefficient data to decode the digital audio signal. Subsequently, the decoder 7 proceeds to step SP15, executes post-processing, and then records and outputs the digital audio signal DA4 decoded in step SP16 in, for example, a buffer memory or the like.

Subsequently, the decoder 7 proceeds to step SP17 to determine whether or not the processing of the digital audio signal DA3 has been completed. If a negative result is obtained here, the process returns to step SP11, whereas a positive result is obtained. Then, the process moves to step SP18 and ends this processing procedure.

(1-2) Operation of First Embodiment In the above configuration, digital audio signals provided to ordinary households by CS digital broadcasting (FIG. 2)
In each home 4, the data is received and decoded by the IRD 5, and is recorded on the mini disk 6 or the like depending on the user.

At this time, in this embodiment, the digital audio signal DA1 is subjected to a pre-processing corresponding to the format (FIGS. 1 and 3) in a data compression process for recording on the mini-disc 6 or the like. After a process of orthogonal transformation, which is a general process, a predetermined bit of predetermined coefficient data among a plurality of coefficient data obtained as a result is set to a logical value corresponding to the data DC of the user watermark. As a result, the data DC of the user watermark is embedded in the coefficient data, and then the encoding process is performed to complete the data compression process. As a result, in the data compression process, the data DC of the user watermark is embedded in the digital audio signal DA1 and is recorded on the mini disk 6 or the like.

The digital audio signal recorded on the mini-disc 6 in this way is legal when the user individually listens to it and enjoys it, and if placed on a distribution channel in the market, it is a copyright infringement. . In this case, in a recording medium suspected of illegal copying in the market, the digital audio signal DA3 obtained by encoding the coefficient data obtained in the reproducing process of the corresponding reproducing apparatus is supplied to the decoder 7 without being subjected to data expansion processing. Is entered.

In the digital audio signal DA3 (FIGS. 3 and 4), after the coefficient data is decoded, user watermark data DC is extracted from a predetermined bit of the predetermined coefficient data, and the coefficient data is subjected to inverse orthogonal transform. After being processed, it is post-processed to reproduce the original digital audio signal. Thereby, in the data decompression process, the data of the user watermark is decoded, and it is possible to detect an illegal copy. Further, by detecting the data of the user watermark, it is possible to prohibit the creation of a secondary copy, thereby preventing further unauthorized copying.

(1-3) Effects of the First Embodiment According to the above configuration, the data of the user watermark is embedded in the coefficient data obtained by the orthogonal transformation, so that the data of the user watermark is compressed in the data compression process. Data can be embedded.

(2) Second Embodiment FIG. 6 is a block diagram showing an encoder according to a second embodiment of the present invention. This encoder 30 is applied to the case where the digital audio signal DA1 is recorded on the mini disk 6 instead of the encoder 10 described above with reference to FIG. 1, and compresses the digital audio signal DA1 by ATRAC.

In this encoder 30, the preprocessing unit 3
1 pre-processes the digital audio signal DA1 by dividing the band of the digital audio signal DA1. That is, in the pre-processing unit 31, the band dividing unit 32
Is composed of, for example, a QMF (Quadrature Mirror Filter) filter, and divides the digital audio signal DA1 into a digital audio signal having a frequency band of 0 to 11 [kHz] and a digital audio signal having a frequency band of 11 to 22 [kHz].

The subsequent band division unit 33 is constituted by, for example, a QMF (Quadrature Mirror Filter) filter, like the band division unit 32, and converts a digital audio signal of a frequency band of 0 to 11 [kHz] output from the band division unit 32. The digital audio signal is divided into a digital audio signal having a frequency of 0 to 5.5 [kHz] and a digital audio signal having a frequency of 5.5 to 11 [kHz].

The MDCT calculation units 34, 35 and 36 respectively output the frequencies 0 to 5.5 [k
[Hz] band, a frequency of 5.5 to 11 [kHz], and a frequency of 11 to 22 [kHz] are orthogonally transformed to output coefficient data SC. At this time, the MDCT calculation units 34, 35, and 36 orthogonally transform the digital audio signal by the MDCT conversion process, and output coefficient data SC based on frequency spectrum data indicating the signal level of each frequency spectrum.

The encoding section 37 sets a predetermined bit of the predetermined frequency spectrum data SC in the frequency spectrum data SC thus obtained to a logical value corresponding to the data DC of the user watermark. The data DC of the user watermark of the frequency spectrum data is embedded.

The adaptive coding section 38 performs adaptive coding processing on the output data SC1 of the encoding section 37 and outputs the result. Thereby, the encoder 30 embeds the data DC of the user watermark in the digital audio signal DA1 in the data compression process by ATRAC.

FIG. 7 is a flowchart showing the processing procedure of the encoder 30. The encoder 30 proceeds from step SP31 to step SP32, and reads the digital audio signal DA1 based on the time-series sample data. Subsequently, the encoder 30 proceeds to step SP33
In step SP34, the digital audio signal DA1 is divided into bands.
Perform orthogonal transformation processing by transformation processing.

At step SP35, the encoder 30 embeds the user watermark data DC into the frequency spectrum data SC obtained by the orthogonal transformation process, and then at step SP36, executes the adaptive coding process. The data resulting from the encoding process is recorded in, for example, a buffer memory and output.

The encoder 30 proceeds to the next step SP38.
In, it is determined whether or not the data of the digital audio signal DA1 is completed. If a negative result is obtained here,
While returning to step SP32, if a positive result is obtained, the procedure moves to step SP39, and this processing procedure ends.

In this embodiment, when decoding the digital audio signal recorded by compressing the data in this manner, the frequency spectrum obtained by executing the processing corresponding to FIGS. 6 and 7 is performed. Data D of user watermark from data SC
C will be detected.

According to the configuration shown in FIGS. 6 and 7, the ATR
In the data compression process by the AC, the data DC of the user watermark is converted to the digital audio signal DA.
1 can be embedded.

(3) Third Embodiment FIG. 8 is a block diagram showing an encoder according to a third embodiment of the present invention. This encoder 40 is applied to the case where the digital audio signal DA1 is recorded on the mini disk 6 instead of the encoder 10 described above with reference to FIG. 1, and compresses the digital audio signal DA1 by ATRAC2.

In the encoder 40, the preprocessing unit 4
1 pre-processes the digital audio signal DA1 by dividing the band of the digital audio signal DA1 and correcting the signal level. That is, the preprocessing unit 4
1, the band dividing unit 42 is, for example, a PQF (Polyph
ase Quadrature Filter), and converts the digital audio signal DA1 to a frequency of 0 to 5.5 [k
Hz] band, frequency 5.5 to 11 [kHz] band, frequency 11 to 18.5 [kHz] band, frequency 18.5 to 2
The digital audio signal is divided into 2 [kHz] band digital audio signals.

The gain correction units 43, 44, 45, 46
Each of the frequency divisions 0 to 5.5 [kHz]
z] band, frequency 5.5 to 11 [kHz] band, frequency 11 to 18.5 [kHz] band, frequency 18.5 to 22
It receives digital audio signals in the [kHz] band, corrects the signal levels, and outputs the signals.

The orthogonal transform unit 47 performs an orthogonal transform process on the digital audio signal output from the preprocessing unit 41 by an MDCT process. That is, in the orthogonal transform unit 47, the MDCT calculation units 48, 49, 50, and 51 receive the digital audio signals from the gain correction units 43, 44, 45, and 46, respectively, and output frequency spectrum data SC by MDCT conversion processing.

The selection circuit 52 receives the frequency spectrum data SC from the orthogonal transform unit 47, and determines whether or not predetermined frequency spectrum data SC of the frequency spectrum data SC has a tone characteristic or a noise characteristic. Is output to the encoding unit 54 instead of the adaptive encoding unit 53. Here, the tone property refers to a case where the frequency spectrum data SC as the target is continuous at a predetermined value or more. On the other hand, the noise property refers to a case where the target frequency spectrum data SC is not continuous at a predetermined value or more.

The encoding section 54 controls the frequency spectrum data SC input from the selection circuit 52 in this manner.
By setting a predetermined bit of the predetermined frequency spectrum data SC to a logical value corresponding to the data DC of the user watermark, the data DC of the user watermark of the frequency spectrum data SC is embedded and output.

The adaptive coding section 53 performs adaptive coding processing on the frequency spectrum data SC output from the selection circuit 52 and the encoding section 54 and outputs the result. Thus, the encoder 40 embeds the data of the user watermark in the digital audio signal DA1 in the data compression process by ATRAC2.

FIG. 9 is a flowchart showing the processing procedure of the encoder 40. The encoder 40 proceeds from step SP41 to step SP42, and reads the digital audio signal DA1 based on the time-series sample data. Subsequently, the encoder 40 proceeds to step SP43.
After the digital audio signal DA1 is divided into bands, the signal level is corrected in the following step SP44.

The encoder 40 proceeds to the next step SP45.
, An orthogonal transform process by the MDCT transform process is executed, and in a succeeding step SP46, it is determined whether or not the predetermined frequency spectrum data SC of the frequency spectrum data SC obtained by the orthogonal transform process has a tone characteristic.
Here, when a positive result is obtained, the encoder 40 embeds the user watermark data in the frequency spectrum data SC obtained by the orthogonal transformation process in step SP47, and then proceeds to step SP48. On the other hand, if a negative result is obtained, step SP
The process directly proceeds from step 46 to step SP48.

The encoder 40 determines in this step SP48
In step, an adaptive encoding process is performed, and
In P49, the data resulting from the encoding process is recorded in, for example, a buffer memory and output. Encoder 40
Determines in step SP50 whether or not the data of the digital audio signal DA1 is completed. If a negative result is obtained here, the process returns to step SP42, whereas if a positive result is obtained, the process proceeds to step SP51. The leverage procedure ends.

In this embodiment, when decoding the digital audio signal recorded by compressing the data in this manner, the frequency spectrum obtained by executing the processing corresponding to FIGS. 8 and 9 is performed. Data D of user watermark from data SC
C will be detected.

According to the configuration shown in FIGS. 8 and 9, the ATR
In the data compression process by AC2, user watermark data DC is converted to digital audio signal D.
It can be embedded in A1.

(4) Fourth Embodiment FIG. 10 is a block diagram showing an encoding unit applied to an encoder according to a fourth embodiment. This encoding unit 60 is applied in place of the encoding units 37 and 54 described in the second and third embodiments.

That is, in the calculation result obtained by the MDCT conversion, the spectrum value changes according to the phase of the input signal. Therefore, in the configurations according to the above-described second and third embodiments, when the user watermark data is embedded in predetermined bits of the frequency spectrum data, and when the digital watermark is copied in the form of a digital signal, the data is copied from the copied recording medium. While user watermark data can be detected correctly, if the data is once converted to an analog signal format and copied, the frequency spectrum data in which the user watermark data DC has been embedded correctly according to the phase of the analog-to-digital conversion process. It may not be detected.

This makes it difficult to find the data DC of the user watermark and improves the confidentiality of the data. On the other hand, when embedding the data DC of the user watermark by finely setting the amplitude of the frequency spectrum data, DC itself becomes difficult to discover.

However, when looking at the change in the sampling phase of a single sine wave input signal with respect to two adjacent spectrum values (spectral values obtained by MDCT conversion), the phase is shifted by π / 2, and the phase of the input signal is shifted. It turns out that it changes trigonometrically according to. That is, it was shown that in these two spectra, the RMS (root mean square) of the spectrum value takes a constant value without depending on the phase.

According to the MDCT calculation result, when the input signal is a single sine wave, even if the phase is changed, the distribution of the spectrum energy is concentrated around four peak spectrum values, which are almost theoretical values. these frequency spectrum value spec _i the spectrum value is intended to focus the energy, one frequency spectrum values spec _i-1 of the immediately preceding, a spectrum value _{spec i + 1, spec i +} 2 2 pieces of the immediately following It turned out to be.

Thus, in this embodiment, the representative value calculating section 61 calculates the frequency spectrum of the frequency 3 [kHz] as spectrum _i , the immediately preceding frequency spectrum value as spectrum _i−1 , and the two immediately following frequency spectrum as spectrum _i−1 . The spectrum value of
The arithmetic processing of the following equation is executed with ped _{i + 1} and ped _{i + 2} . Accordingly, the representative value calculation unit 61 adds the square of the absolute value sum of the even-numbered spectrum values and the square of the absolute value sum of the odd-numbered spectrum values, calculates the square root of the added value, and calculates Representative value P of the frequency spectrum value of
Is calculated.

[0065]

(Equation 1)

The representative value calculation section 61 outputs the representative value P calculated in this way and the frequency spectrum data to the subsequent bit operation section 62.

[0067] Bit manipulation unit 62, thus to determine the representative value P obtained for each sample block is whether or not the value 2 ⁿ or more, if more than the value 2 ^n, user watermark to the representative value P The data DC is embedded. In the embedding process, the bit operation unit 62 sets, for example, the fourth bit higher than the least significant bit to a logical value corresponding to the polarity of the input data DC. Note that sound quality degradation can be suppressed as the bit for setting this logic level is closer to the most significant bit, but the resistance to attack becomes weaker. Therefore, the bit for setting this logic level is appropriately selected as needed. In this embodiment, in order to make the sound quality deterioration difficult to perceive, a bit operation process is selectively executed on a sample block whose representative value P is equal to or ^larger than 2 ⁿ .

The bit operation unit 62 converts the digital audio signal in which the user watermark data DC is embedded into an analog signal, and then converts the digital audio signal into a digital signal again. An offset value is set below the bit in which the logical value is set so that the logical value can be correctly reproduced in the bit in which the logical value is set in accordance with the quantization error.

That is, such a quantization error can be examined by, for example, addition and subtraction processing on the least significant bit.
In the process of adding the value 1 to the least significant bit, no change in the logical value is observed. However, in the process of subtracting the value 1 from the least significant bit, a change in the logical value is observed.

Thus, the bit operation unit 62 sets the logical value of a bit lower than the predetermined bit so that the logical value of the user watermark data DC does not change at least by adding and subtracting the least significant bit. And the representative value P is correctly determined by the quantization error.
Enables detection of user watermark data even when cannot be played.

The bit operation section 62 outputs the correction data P1 of the representative value P obtained by performing the bit operation to the spectrum correction section 63 together with the frequency spectrum data SC.

The spectrum correcting section 63 corrects the frequency spectrum value used for calculating the representative value P based on the representative value P1 subjected to the bit operation as described above. Here, the spectrum correction unit 63 executes the following arithmetic processing, thereby proportionally distributing the change of the representative value P due to the bit operation according to each spectrum value spec.

[0073]

(Equation 2)

Here, "spec" indicates the representative value P before the bit operation, and "spec '" indicates the representative value P after the bit operation. As a result, the encoding unit 60 determines that, in the representative value P which is the sum of squares of a plurality of spectrum values in the vicinity of the predetermined frequency spectrum of the frequency spectrum signal, the predetermined bit is the data D of the user watermark.
The input data DC is embedded in the frequency spectrum data SC1 so as to have a logical value corresponding to the logical value of C.

Further, at this time, by setting the predetermined bit of the representative value P to a logical value corresponding to the logical value of the data DC of the user watermark, the data is dispersed in a narrow frequency band near the predetermined frequency spectrum and the user watermark is set. The mark data DC is embedded.

FIG. 11 is a flowchart showing a processing procedure in the encoding section 60. Encoding unit 6
0 moves from step SP61 to step SP62,
After calculating the representative value P, the user watermark data DC is embedded in the representative value P in step SP63. At this time, the encoding unit 60 calculates the representative value P
Is determined to be greater than or equal to a predetermined value.
The user watermark data DC is embedded.

Subsequently, the encoding unit 60 proceeds to step SP
64, the user watermark data DC
Is set to logic 0, and a predetermined offset value is further added, so that the user watermark data DC can be correctly detected even by a quantization error.

Subsequently, the encoding unit 60 proceeds to step SP
The process proceeds to 65 where the frequency spectrum value used for detection of the representative value P is corrected and output, and this processing procedure is terminated in the subsequent step SP66.

FIG. 12 is a block diagram showing a decoding unit corresponding to the encoding unit 60 of FIG. The decoding unit 70 is applied to a decoder that expands data when the user watermark data DC is embedded in the frequency spectrum data by the encoding unit in FIG. Further, the present invention is applied to a case where data DC of a user watermark is detected from a digital audio signal whose data has been expanded.

That is, in the data decompression process, the decoding unit 70 decodes the decoded frequency spectrum data S
Input coefficient data by C1. When detecting the data DC of the user watermark from the digital audio signal whose data has been decompressed, coefficient data based on the frequency spectrum data SC1 generated by the MDCT conversion processing is input.

In the decoding section 70, the representative value calculating section 7
1 reads the frequency spectrum data SC1 for each sample block. Further, the representative value calculation unit 71 executes the arithmetic processing of Expression (1) for each sample block,
Thus, a representative value P1 of a predetermined frequency is detected.

When the representative value P1 is equal to or greater than the value corresponding to the predetermined value determined by the encoding unit 60, the bit detector 72 selectively takes in the predetermined bits of the representative value P1. The bit detection unit 72 outputs the data of the bit string sequentially taken in from the sample blocks in this way as the data DC of the user watermark.

FIG. 13 is a flowchart showing a processing procedure of the decoding unit 70. The decoding unit 70 proceeds from step SP71 to step SP72, where the representative value P1
Is calculated, in step SP73, the representative value P1
, And the user watermark data DC is detected. At this time, the decoding unit 70 determines whether or not the representative value P1 is equal to or larger than a predetermined value, and only when the representative value P1 is equal to or larger than the predetermined value, fetches predetermined bits of the representative value P1 to detect the data DC of the user watermark.

When the user watermark data DC is detected in this manner, the decoding unit 70 proceeds to step SP74 and ends this processing procedure.

According to the configuration shown in FIGS.
When embedding the data of the user watermark in the frequency spectrum data by the CT conversion, the data DC of the user watermark can be reliably detected even if the data is once converted into an analog signal.

(5) Other Embodiments In the above-described fourth embodiment, the case where the representative value is calculated by calculating the square root of the sum of squares calculated from the four spectrum values has been described. However, the present invention is not limited to this, and the sum of squares may be used as a representative value in a practically sufficient range. By doing so, the overall configuration can be simplified accordingly.

In the above-described fourth embodiment,
A case has been described in which the representative value is calculated by adding the square value of the sum of absolute values of the even-numbered spectrum values and the square value of the sum of absolute values of the odd-numbered spectrum values from the four spectrum values. However, the present invention is not limited to this, and if sufficient accuracy for practical use can be ensured, a representative value may be calculated based on a target frequency spectrum value and an adjacent spectrum value. A representative value may be calculated from four or more spectrum values. Instead of the representative value, various data may be embedded in each frequency spectrum data.

In the above-described embodiment, the case where desired data is embedded in coefficient data and representative values by orthogonal transformation has been described. However, the present invention is not limited to this. For example, after desired data is subjected to PSK modulation, Widely applicable when embedding desired data by variously operating coefficient data by orthogonal transformation, such as when generating coefficient data with a predetermined frequency spectrum by spreading with random number data and adding this coefficient data to the corresponding coefficient data can do.

Further, in the above-described embodiment, the case where the data of the user watermark is embedded has been described. However, the present invention is not limited to this, and is widely applied to a case where various information is superimposed and transmitted as necessary. can do.

In the above-described embodiment, the case where a digital audio signal is recorded on an optical disk and reproduced is described. However, the present invention is not limited to this case. For example, when various information is transmitted via the Internet. Etc. can be widely applied.

In the above-described embodiment, the digital audio signal is converted into frequency spectrum data by MDCT, which is one of orthogonal transforms, and the frequency spectrum data is converted into digital audio data by inverse MDCT. Although the case of converting to a signal has been described, the present invention is not limited to this.
Dio, Dolby AC2, Dolby AC3, T
It can be widely applied to data compression processing of various audio signals by orthogonal transformation processing such as win VQ, and furthermore, processing by various orthogonal transformations such as discrete cosine transformation processing and Haar transformation processing applied to video signals Can be widely applied to.

[0092]

As described above, according to the present invention, by embedding input data in coefficient data obtained by orthogonal transformation, a signal processing apparatus capable of embedding user watermark data and the like in a data compression process. A signal processing method, a signal processing device, and a recording medium using the signal processing method can be obtained.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an encoder according to a first embodiment of the present invention.

FIG. 2 is a schematic diagram illustrating a system for detecting unauthorized copying to which the encoder of FIG. 1 is applied.

FIG. 3 is a flowchart showing a processing procedure of the encoder of FIG. 1;

FIG. 4 is a block diagram illustrating a decoder applied to the system of FIG. 2;

FIG. 5 is a flowchart showing a processing procedure of the decoder of FIG. 4;

FIG. 6 is a block diagram illustrating an encoder according to a second embodiment.

FIG. 7 is a flowchart showing a processing procedure of the encoder of FIG. 6;

FIG. 8 is a block diagram illustrating an encoder according to a third embodiment.

FIG. 9 is a flowchart illustrating a processing procedure of the encoder of FIG. 8;

FIG. 10 is a block diagram illustrating an encoding unit applied to an encoder according to a fourth embodiment.

FIG. 11 is a flowchart illustrating a processing procedure of an encoding unit in FIG. 10;

FIG. 12 is a block diagram illustrating a decoding unit applied to a decoder according to a fourth embodiment.

FIG. 13 is a flowchart illustrating a processing procedure of a decoding unit in FIG. 12;

[Explanation of symbols]

6 ... mini disk, 7 ... decoder, 10, 30, 4
0 ... encoder, 11, 31, 41 ... pre-processing unit, 1
2, 47 ... orthogonal transform unit, 13, 37, 54, 60 ...
Encoder unit, 22, 70... Decoding unit, 23... Inverse orthogonal transform unit, 24.

Claims (19)

[Claims] 1. An orthogonal transformation means for orthogonally transforming an input signal to generate coefficient data, and a coefficient data processing for embedding input data in predetermined coefficient data of the coefficient data to generate correction data of the coefficient data. And a coding means for coding correction data of the coefficient data. 2. The input signal is an audio signal, wherein the orthogonal transform unit performs an improved discrete cosine transform process on the input signal to generate the coefficient data based on frequency spectrum data, and the encoding unit includes: 2. The signal processing apparatus according to claim 1, wherein the coefficient data based on the frequency spectrum data is adaptively encoded. 3. The input signal is an audio signal, the signal processing device performs signal processing on the audio signal by a predetermined preprocessing unit, and inputs the audio signal to the orthogonal transform unit. The audio signal is band-divided and output, and the orthogonal transform unit executes an orthogonal transform process by an improved discrete cosine transform process for each audio signal of each band output from the pre-processing unit. 2. The signal processing apparatus according to claim 1, wherein coefficient data based on frequency spectrum data is adaptively encoded. 4. The input signal is an audio signal, the signal processing device performs signal processing on the audio signal by a predetermined preprocessing unit, and inputs the audio signal to the orthogonal transform unit. The audio signal is band-divided and the signal level is corrected and output, and the orthogonal transform unit executes an orthogonal transform process by an improved discrete cosine transform process for each audio signal of each band output from the preprocessing unit. The signal processing apparatus according to claim 1, wherein the encoding unit performs adaptive encoding processing on coefficient data based on frequency spectrum data. 5. The coefficient data processing means, when the predetermined coefficient data is continuously equal to or more than a predetermined value,
The signal processing device according to claim 1, wherein the input data is embedded. 6. A decoding unit that decodes an input signal and outputs coefficient data, a coefficient data processing unit that detects predetermined data embedded in the coefficient data, and performs an inverse orthogonal transform process on the coefficient data and outputs the result. A signal processing device comprising: 7. The signal processing apparatus according to claim 6, wherein the inverse orthogonal transform unit performs an inverse improved discrete cosine transform process on the coefficient data and outputs an audio signal. 8. The correction data of the coefficient data is generated by embedding the input data in predetermined coefficient data of the coefficient data generated by performing an orthogonal transformation process on the input signal, and then correcting the correction data of the coefficient data. A recording medium characterized by being converted and recorded. 9. The input signal is an audio signal, the orthogonal transform process is an improved discrete cosine transform process that generates the coefficient data based on frequency spectrum data from the input signal, and the encoding process is 9. The recording medium according to claim 8, wherein the coefficient data is adaptively encoded using frequency spectrum data. 10. The input signal is an audio signal, wherein the orthogonal transformation process performs signal processing on the audio signal by a predetermined pre-processing to perform the orthogonal transformation processing, and the pre-processing converts the audio signal into a band. The orthogonal transform process is a process of performing an improved discrete cosine transform process for each audio signal of each band by the pre-process, and the encoding process is a process of adaptively encoding the coefficient data. 9. The recording medium according to claim 8, wherein: 11. The input signal is an audio signal, wherein the orthogonal transform processing performs signal processing on the audio signal by predetermined pre-processing to perform the orthogonal transform processing. Dividing and correcting the signal level, the orthogonal transform process is a process of performing an improved discrete cosine transform process for each audio signal of each band by the pre-process, and the encoding process is an adaptive encoding of the coefficient data. 9. The recording medium according to claim 8, wherein the recording medium is a process for performing the following. 12. The recording medium according to claim 8, wherein the input data is embedded when the predetermined coefficient data continuously exceeds a predetermined value. 13. An input signal is subjected to an orthogonal transformation process to generate coefficient data, input data is embedded in predetermined coefficient data of the coefficient data to generate correction data of the coefficient data, and correction of the coefficient data is performed. A signal processing method comprising encoding data. 14. The input signal is an audio signal, wherein the orthogonal transform process performs an improved discrete cosine transform process on the input signal to generate the coefficient data based on frequency spectrum data, and the encoding process includes: 14. The signal processing method according to claim 13, wherein coefficient data based on frequency spectrum data is adaptively encoded. 15. The input signal is an audio signal, wherein the signal processing method performs signal processing on the audio signal by predetermined pre-processing to perform the orthogonal transform processing, and the pre-processing includes: The orthogonal transform processing is performed for each audio signal of each band by the pre-processing, and is an orthogonal transform processing by an improved discrete cosine transform processing.The encoding processing adapts coefficient data based on frequency spectrum data. 14. The signal processing method according to claim 13, wherein encoding processing is performed. 16. The input signal is an audio signal, wherein the signal processing method performs signal processing on the audio signal by predetermined pre-processing to perform the orthogonal transform processing, and the pre-processing includes: Dividing and correcting the signal level, the orthogonal transform process is an orthogonal transform process by an improved discrete cosine transform process for each audio signal of each band by the signal process, and the encoding process is a coefficient by frequency spectrum data. 14. The signal processing method according to claim 13, wherein data is adaptively encoded. 17. The signal processing method according to claim 13, wherein the input data is embedded when the predetermined coefficient data continuously exceeds a predetermined value. 18. A signal which decodes an input signal to generate coefficient data, detects predetermined data embedded in the coefficient data, and outputs the coefficient data by performing an inverse orthogonal transform process. Processing method. 19. The method according to claim 1, wherein the inverse orthogonal transform process is a process of performing an inverse improved discrete cosine transform on the coefficient data and outputting an audio signal.
9. The signal processing method according to 8.