JP2001282235A - Device and method for embedding watermark information and device and method for reading watermark information - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a watermark information embedding device which can embed watermark data without depending on the time information of an MIDI event. SOLUTION: A watermark information embedding device 1 includes an object event extracting means 2 which extracts an object event having a specific parameter from original MIDI data, a target histogram computing means 3 which computes a target histogram on the basis of the object event number extracted by the means 2 and the watermark data, a histogram computing means 4 which classifies the object event extracted by the means 3 on the basis of the parameter and computes a histogram and a parameter changing means 5 which compares the histogram computed by the means 4 and the target histogram and changes the value of the parameter of the object event when the histogram and the target histogram are different.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to MIDI data used as performance information of a communication karaoke or the like.
In order to prevent unauthorized copying or use of MIDI data, information about the creator such as copyright is used as watermark information (di-
The present invention relates to a watermark information embedding device and a watermark information embedding method for recording as a watermark, and a watermark information reading device and a watermark information reading method for reading the watermark information.

[0002]

2. Description of the Related Art In recent years, content data such as image data and audio data has been digitized, distributed and recorded on a recording medium such as a CD-ROM, and transmitted and received via a network such as the Internet. As a result, illegal use of the content data, such as unauthorized copying or sale of the copied data to the copyright holder without permission, has become a problem.

In particular, since MIDI data used as performance information for communication karaoke and the like is used as music information itself, a copyright is generated for each MIDI data, and it is necessary to protect the copyright. is there. Also,
Recently, as a communication karaoke not only for business use but also for home use, a software synthesizer for transmitting music data (MIDI data) via the Internet and operating the music data on a home personal computer has appeared. . In such a home-use communication karaoke, music data can be stored on a hard disk or the like so that music data (MIDI data) can be used repeatedly as many times as it is received.

Therefore, music data (MIDI data)
Is illegally copied, and the copied data can be used commercially without the permission of the right holder.

[0005] Therefore, as a means for preventing such unauthorized use of MIDI data, a digital watermark embedding method for recording information about a creator such as copyright in content data as watermark information (digital watermark). Is being researched and developed.

[0006] The conditions necessary for this digital watermark include:
(1) embedding watermark information so that the content data will be broken if it is forcibly removed (watermark information indicating copyright remains in the content data); and (2) where watermark information is embedded in the content data. (3) Content data remains in its original state even when watermark information is embedded.

[0007] As a digital watermarking technique that satisfies these conditions, a conventional system for distributing MIDI data in communication karaoke and the like focuses on note events and control events included in the MIDI data, and a part of the data of these events is used. Digital watermarking is performed by replacing the data with other data so as not to degrade the music quality. In this method, the order of embedding the watermark data is determined based on the time information of the MIDI event.

Further, an error correction technique such as an RS code is used to correct an error so that the watermark data is hardly erased.

[0009]

SUMMARY OF THE INVENTION However, MID
There may be a plurality of I events having exactly the same time information. In such a case, in order to determine the order of embedding the watermark data, it is conceivable to take measures such as deciding the order of embedding the information other than the time information together with the time information. However, if the time information is intentionally changed, the MIDI event having the exact same time information will have different time information, and the criteria for determining the order of embedding the watermark data will change. Therefore, there has been a problem that watermark data cannot be read from MIDI events in an expected order.

In the case where the MIDI event is lost, the data reading is synchronized.
There is also a problem that the error correction does not function effectively and the watermark data cannot be read.

SUMMARY OF THE INVENTION The present invention has been made in view of the above circumstances, and has as its object to provide a watermark information embedding apparatus and a watermark information embedding method capable of embedding watermark data without depending on time information of a MIDI event. An object of the present invention is to provide a watermark information reading device and a watermark information reading method for reading watermark information.

[0012]

According to a first aspect of the present invention, there is provided a watermark information embedding apparatus for embedding watermark data in original MIDI data. A target event extracting unit for extracting a target event having a specific parameter from the original MIDI data; a target histogram is calculated based on the number of the target events extracted by the target event extracting unit and the watermark data. Target histogram calculation means,
A histogram calculation unit that classifies the target event extracted by the target event extraction unit based on the parameter to calculate a histogram, and compares the histogram calculated by the histogram calculation unit with the target histogram, And a parameter changing means for changing a value of the parameter of the target event when different.

According to the first aspect of the present invention, the watermark data can be embedded without depending on the time information, and an error may occur in the MIDI data with the embedded watermark due to noise or some event may occur. Even if the data is processed or deleted, synchronization of data reading will not be lost, so that errors can be suppressed by ordinary error correction technology.

A watermark information embedding method according to a second aspect of the present invention is a watermark information embedding method for embedding watermark data in original MIDI data,
A target event extraction step of extracting a target event having a specific parameter from the original MIDI data, a number of the target events extracted in the target event extraction step, and the watermark data,
A target histogram calculating step of calculating a target histogram, a histogram calculating step of classifying the target event extracted in the target event extracting step based on the parameter to calculate a histogram, and the histogram calculated in the histogram calculating step. A step of comparing the histogram with the target histogram and, when different, changing a parameter value of the target event.

According to the second aspect of the present invention, the watermark data can be embedded without depending on the time information, and an error may occur in the MIDI data with the embedded watermark due to noise or some event may occur. Even if the data is processed or deleted, synchronization of data reading will not be lost, so that errors can be suppressed by ordinary error correction technology.

A watermark information reading device according to a third aspect of the present invention is a watermark information reading device for reading watermark data embedded in MIDI data, and extracts a target event having a specific parameter from the MIDI data. A target event extracting unit, a histogram calculating unit that classifies the target event extracted by the target event extracting unit based on the parameter to calculate a histogram, and based on the histogram calculated by the histogram calculating unit, Watermark data reading means for reading watermark data.

According to the third aspect of the present invention, the watermark information embedded by the above-described watermark information embedding device can be read.

According to a fourth aspect of the present invention, there is provided a watermark information reading method for reading watermark data embedded in MIDI data, wherein a target event having a specific parameter is extracted from the MIDI data. A target event extraction step, a histogram calculation step of classifying the target event extracted in the target event extraction step based on the parameter to calculate a histogram, and based on the histogram calculated in the histogram calculation step,
Reading a watermark data.

According to the fourth aspect of the present invention, the watermark information embedded by the above-described watermark information embedding method can be read.

[0020]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of a watermark information embedding device according to the present invention will be described below with reference to the drawings.

As shown in FIG. 1, a watermark information embedding device 1 of the present embodiment extracts a target event having a specific parameter from original MIDI data, and a target event extracting unit 2 for extracting the target event. The target histogram calculating means 3 for calculating a target histogram based on the number of target events thus obtained and the watermark data, and the target events extracted by the target event extracting means 2 are classified based on specific parameters to generate a histogram. Histogram calculation means 4 to be calculated
And a parameter change unit 5 that compares the histogram calculated by the histogram calculation unit 4 with the target histogram and, when different, changes the parameter value of the target event.

Next, a process of embedding watermark information by the watermark information embedding device of the present embodiment will be described with reference to the flowchart of FIG.

As shown in FIG. 2, when original MIDI data in which watermark information is embedded is input,
A target event into which watermark information is to be embedded is extracted from the original MIDI data (S201). M
Since the IDI data event includes various events such as a note-on event, a note-off event, a pitch bend event, a program change event, and an expression event, any event may be used as the target event.

Next, when the target event is extracted, the number of events of the extracted target event is calculated (S2).
02). Then, a target histogram is calculated based on the calculated number of events and the watermark data (S203).

FIG. 3 shows an example of the target histogram.
Shown in In the target histogram of FIG.
[N] is set to an average value representing a threshold level, and the remaining H [0] to H [N-1] represent values of watermark data to be embedded. That is, H
The values of [0] to H [N-1] are compared with the values of H [N], respectively. When the value is larger than H [N], data "1" is represented, and when smaller, data "0" is represented. Thus, the values of the watermark data are represented by the values of H [0] to H [N-1].

For example, the histogram shown in FIG. 3 shows a histogram when watermark data (0110) ₂ (binary number) is embedded. Since the watermark data is 4 bits, N = 4. Therefore, H [1] and H [2] larger than the value of H [4] represent data “1”,
H [0] and H [3] smaller than the value of [4] represent data “0”, and the histogram of FIG. 3 represents 4-bit watermark data (0110) ₂ .

Next, a method of calculating the maximum value HL and the threshold level Th in the target histogram will be described.

First, assuming that the number of "1" in the watermark data is Nh and the number of target events calculated in step S202 is Ne, the maximum value HL can be calculated by HL = Ne / (Nh + 0.5). . For example, the number of target events Ne = 250, 4-bit watermark data is (0110)
_In the case of ₂ , the number of "1" is Nh = 2, so that HL = 250 / (2 + 0.5) = 100. Also, the threshold level Th
Can be obtained by Th = HL / 2, so that Th = 100/2 = 50 can be obtained.

In this way, the target histogram representing the watermark data can be calculated.

Next, the target events extracted in step S201 are classified based on specific parameters, and a histogram of the original MIDI data is calculated (S201).
204).

First, paying attention to a specific parameter of the extracted target event, the remainder obtained by dividing the value of the parameter by N + 1 when the number of bits of the watermark data is N is Ind.
Classify as ex. This Index is represented by Index = (MIDI event parameter)% (N + 1). Then, the number of target events for which Index is equal is calculated, and values of H [0] to H [N] are obtained. For example, when attention is paid to the velocity of the note-on event as a parameter, since the velocity is represented by 7 bits, it has a value from 0 to 127. When the number of bits of the watermark data is 4, the remainder obtained by dividing the value by 5 (= N + 1) is classified as Index.
Calculate the number of target events that have the same remainder and calculate H [0]
To the value of H [N].

FIG. 4 shows an example of a histogram when N = 4. In the histogram shown in FIG. 4, the number of target events H [0] having a remainder of 0 is 68, the number of target events H [1] having a remainder of 1 is 50,..., The target having a remainder of 4 This shows that the number of events H [4] is 28.

After the histogram and the target histogram are calculated in this way, these histograms are compared to determine whether they are equal (S205).

If not equal, the parameter value of the target event is changed so that the histogram approaches the target histogram (S206).

For example, in the histogram of FIG.
The parameter, for example, the velocity value is changed so that the value H [0] of the histogram 0 approaches the value of H [0] = 0 in FIG.

When the value of the parameter is changed,
Returning to step S204, the histogram is calculated again, and comparison with the target histogram is performed in step S205. This process is repeated until the histogram becomes equal to the target histogram, and when the histogram becomes equal, the watermark information embedding process ends.

As shown in FIG. 5, H [0] to H [0]
Even if the change amount of [N-1] is reduced, the threshold level
4 bits of watermark data (0
110) ₂Can be embedded. In this case,
Embed watermark data because meter change is small
Less degradation of the original music quality
Can be done.

As described above, the watermark data can be embedded in the parameter by changing the parameter to change the histogram to the target histogram. Therefore, the watermark data can be embedded without depending on the time information.

Also, even if an error occurs due to noise in the MIDI data in which the watermark is embedded, or even if some events are processed or deleted, the synchronization of the data reading is not disrupted. Error propagation can be suppressed by the correction technique.

Next, an embodiment of a watermark information reading device according to the present invention will be described with reference to the drawings.

As shown in FIG. 6, the watermark information reading device 1 of this embodiment extracts a target event having a specific parameter from MIDI data, and the target event extracting unit 62 extracts the target event. A histogram calculating unit 63 for classifying the target events based on the parameters to calculate a histogram, and a watermark data reading unit 64 for reading watermark data based on the histogram calculated by the histogram calculating unit 63 are included.

Next, a process of reading watermark information by the watermark information reading device of the present embodiment will be described.

First, the MI in which the watermark information is embedded
When DI data is input, a target event in which watermark information is embedded is extracted from the MIDI data.

When the target event is extracted, a histogram of the MIDI data is calculated based on the parameters of the target event. The calculation of this histogram is
Since the same calculation as the calculation of the histogram performed in the watermark information embedding process is performed, detailed description is omitted.

In the calculation of the histogram, the histogram is calculated in the same manner as in step S204 of FIG. 2, and the histogram shown in FIG. 3 is obtained as the calculation result.

Therefore, the values of H [0] to H [N-1] are compared with the values of H [N], and H [0] to H [N-1] are obtained in the same manner as when watermark data is embedded. ] And H [N]
The watermark data can be read according to the magnitude of the value of the watermark data.

For example, in FIG. 3, the value is larger than the value of H [4].
H [1] and H [2] indicate the data “1”,
H [0] and H [3] smaller than the value of H [4] are data
It represents "0". Therefore, embedded
The watermark data is (0110) ₂Can be read as
it can.

Thus, by embedding the watermark data by classifying the target event based on the parameters and calculating the histogram.

[0049]

As described above, according to the watermark information embedding apparatus and method of the present invention, the watermark data can be embedded without depending on the time information, and the watermark embedded MIDI data can be embedded in the MIDI data. Even if an error occurs due to noise, or even if some events are processed or deleted, no error occurs in synchronization in data reading, so that error propagation can be suppressed by ordinary error correction technology. Further, according to the watermark information reading device and method of the present invention, the watermark information embedded by the above-described watermark information embedding device can be read.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of an embodiment of a watermark information embedding device according to the present invention.

FIG. 2 is a flowchart illustrating a process of embedding watermark data by the watermark information embedding device illustrated in FIG. 1;

FIG. 3 is a diagram illustrating an example of a target histogram calculated by the watermark information embedding device illustrated in FIG. 1;

4 is a diagram showing an example of a histogram calculated by the watermark information embedding device shown in FIG.

FIG. 5 is a diagram illustrating an example of a histogram calculated by the watermark information embedding device illustrated in FIG. 1;

FIG. 6 is a block diagram showing a configuration of an embodiment of a watermark information reading device according to the present invention.

[Description of Signs] 1 watermark information embedding device 2, 62 target event extracting means 3 target histogram calculating means 4, 63 histogram calculating means 5 parameter changing means 61 watermark information reading device 64 watermark data reading means

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) G10K 15/04 302 G10K 15/04 302D 9A001 H04N 1/387 H04N 1/387 F term (Reference) 5B057 AA20 CA19 CE08 CH12 DC19 DC32 5C076 AA14 BA06 5D108 BB06 BG06 5D378 QQ02 QQ30 5J104 AA14 AA15 PA14 9A001 EE07 HZ27

Claims (4)

[Claims] 1. A watermark information embedding device for embedding watermark data in original MIDI data, comprising: target event extracting means for extracting a target event having a specific parameter from the original MIDI data; A target histogram calculating unit that calculates a target histogram based on the number of the extracted target events and the watermark data, and classifying the target events extracted by the target event extracting unit based on the parameters. A histogram calculating means for calculating a histogram; and a parameter changing means for comparing the histogram calculated by the histogram calculating means with the target histogram and, when different, changing a parameter value of the target event. Watermark information embedding apparatus characterized by. 2. A watermark information embedding method for embedding watermark data in original MIDI data, comprising: a target event extracting step of extracting a target event having a specific parameter from the original MIDI data; A target histogram calculating step of calculating a target histogram based on the number of the extracted target events and the watermark data, and classifying the target events extracted in the target event extracting step based on the parameters. A histogram calculation step of calculating a histogram, and comparing the histogram calculated in the histogram calculation step with the target histogram, and when different, changing a parameter value of the target event. Watermark information embedding method, characterized in that it comprises a parameter changing step. 3. A watermark information reading device for reading watermark data embedded in MIDI data, wherein the target event extracting means extracts a target event having a specific parameter from the MIDI data, and the target event extracting means extracts the target event. A histogram calculating means for classifying the obtained target event based on the parameter to calculate a histogram; and a watermark data reading means for reading watermark data based on the histogram calculated by the histogram calculating means. Characteristic watermark information reading device. 4. A watermark information reading method for reading watermark data embedded in MIDI data, comprising: a target event extraction step of extracting a target event having a specific parameter from the MIDI data; A histogram calculating step of classifying the obtained target event based on the parameter to calculate a histogram, and a watermark data reading step of reading watermark data based on the histogram calculated in the histogram calculating step. Characteristic watermark information reading method.