KR19990028002A - Moving image copy protection device - Google Patents

Abstract

The present invention relates to a moving picture data copy protection device in which watermark information is inserted only for B frames to prevent illegal copying of digital works. The present invention includes a digital watermark inserter and zigzag scan the quantized discrete cosine transform coefficients into run and level as well as detect the EOB. The address generator generates different addresses depending on the position of the EOB, and the watermark ROM operates only in the case of B frames to read the information designated by the address from among the stored plurality of digital watermark information. The adder inserts the watermark information read from the watermark ROM into the zigzag scanned image data, and the zigzag scanned I and P frames and the B frame into which the watermark information is inserted are multiplexed and variable length encoded. Therefore, the deterioration of the image quality can be minimized by suppressing the increase in the amount of encoded information, and watermark information is not easily removed by users other than the author.

Description

Video data copy protection device

The present invention relates to the prevention of illegal copying of digital works. In particular, when the watermark information is inserted into moving image data, the image quality can be minimized while users other than the author cannot easily remove the watermark information. A moving picture data copy protection device.

In general, in systems such as high-definition television (HD-TV), digital VTR and camcorder, multimedia, video telephone, etc., video and audio are encoded into digital signals and then transmitted or stored on a recording medium, and then decoded and played back. The way to do it is mainly used. In particular, in recent years, research and development for digital video / audio services have been actively conducted worldwide, and efforts have been made for various services in two directions through communication channels as well as one-way information transmission such as conventional television.

Examples include high-definition television, video / audio / games on demand, video telephony, video conferencing, home banking, home shopping, telemedicine, electronic data exchange, electronic transactions, databases, remote lectures, and teleworking. At this time, the information has various forms such as text, figure, voice, sound, still image, video, and general data, and is mainly stored or transmitted in the form of multimedia combined mainly with image and sound. If they are processed by a simple modulation technique, the amount of data becomes excessive, which causes a lot of difficulties in storage or transmission. In particular, the use of a limited communication channel requires a high compression rate, that is, a technique for effectively compressing data in order to maximize data transmission efficiency.

1 is a block diagram showing a schematic configuration of a conventional video data encoding apparatus employing a moving picture compression algorithm recommended by Moving Picture Experts Group 2 (MPEG2). Reference numeral 1 is a first frame memory, and the screen data is configured by storing image data input in units of blocks having a predetermined size in units of frames. A frame / field memory 2 is connected to an output terminal of the first frame memory 1 to separate image data input in units of frames into even field and odd field. The output terminal of the frame / field memory 2 is connected to a subtractor 3, an exercise predictor 13 and an activity calculator 14.

The subtractor 3 subtracts the data fed back from the adaptive predictor 12 from the image data input in field units and outputs error data. The DCT (Discrete Cosine Transform) unit 4 connected to the output of the subtractor 3 converts the input error data into discrete cosine transformed data into frequency domain data. The transform coefficients output from the DCT unit 4 are input to the quantizer 5 and quantized by a predetermined quantization weighting matrix, where the quantizer 5 supplies the quantization level signal from the rate controller 15. According to this signal, the conversion coefficients are converted into representative values of a predetermined level. A variable length encoder and multiplexer 6 and an inverse quantizer 8 are connected to an output terminal of the quantizer 5.

Normally, there are many similar parts between screens, so if there is a slight movement, the motion vector is estimated by calculating the motion, and if the data is compensated using this vector, the difference signal between adjacent screens is very small, so that the transmission data can be further compressed. have. In order to perform such motion compensation, the inverse quantizer 8 performs inverse quantization on the input quantization coefficients corresponding to the quantization described above. The inverse DCT unit 9 connected to the output terminal of the inverse quantizer 8 converts the inverse quantized data into inverse discrete cosine transforms into image data of the spatial domain. The image data of the spatial region output from the inverse DCT unit 9 is input to the adder 10, where it is combined with the data fed back from the adaptive predictor 12 and applied to the second frame memory 11.

The second frame memory 11 reconstructs the screen by storing the input data in units of frames. The second frame memory 11 is connected to the adaptive predictor 12, and the motion predictor through the adaptive predictor 12. It is also connected to (13). The motion predictor 13 connected to the output terminal of the frame / field memory 2 finds data of a pattern most similar to the image data input in units of fields from the data stored in the second frame memory 11 and calculates a motion vector. This motion vector is transmitted to the variable length coding and multiplexer 6 and to the adaptive predictor 12 for use in the decoding apparatus. The adaptive predictor 12 reads data corresponding to the motion vector of the motion predictor 13 from the stored data of the second frame memory 11 and supplies the data to the subtractor 3 and the adder 10.

The variable length encoder and multiplexer 6 connected to the output terminal of the quantizer 5 encodes and compresses the input quantization coefficients by variable length, and compresses the data transmitted by encoding the statistical characteristics of the representative values. The compressed data is multiplexed with other input data, that is, the quantization level signal input from the ratio controller 15 and the motion vectors input from the motion predictor 13. The bit string output from the variable length encoding and multiplexer 6 is applied to the buffer 7 in an irregular state, and the buffer 7 stores the bit string input in the encoded state and transmits it to the receiving side at a constant speed. Or store it on a record carrier.

In addition, the bit string output from the buffer 7 is fed back to the ratio controller 15 and used to adjust the input data amount so that no overflow or underflow occurs. An activity calculator 14 is connected between the ratio controller 15 and the frame / field memory 2, and the activity calculator 14 connected to the output of the frame / field memory 2 is inputted in field units. Find the normalized activity for. The ratio controller 15 connected to the output of the activity calculator 14 calculates the degree of fullness of the buffer using the bit stream fed back from the buffer 7 and calculates the quantization level using this value and the activity. The quantization level signal output from the ratio controller 15 is input to the quantizer 5, and also to the variable length coding and multiplexer 6 to be multiplexed with the encoded data and the motion vector.

As the research and development of the moving picture coding schemes become more active, digitalization of video / audio devices and the explosive spread of the Internet are increasing, and methods of distributing digital works are increasing rapidly. In particular, the use of computers allows anyone to easily copy and edit digital works, so the problem of illegal copying is highlighted. 2. Description of the Related Art Conventionally, the following two technologies are in the spotlight as a countermeasure against illegal copying of digital video / audio data.

The first method is to prevent duplication through encryption and scramble, and to inform the decryption information and the descramble information that can only be operated on the normally purchased video / audio works, and thus the digital video / sound data Reproduction is fundamentally prohibited. The second method is digital watermark technology to prohibit illegal commercial use of digital information, and inserts ID or logo into the video / audio content data in the form of noise so that the user refrains from illegal copying.

However, the first method by encryption and scramble has a problem that can be easily replicated in the case of bit-by-bit or key data leakage. In addition, the second method of inserting watermark information into moving image data has a problem in that an ID or a logo is inserted in the form of noise, thereby reducing coding efficiency and thus degrading image quality.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-mentioned problems, and an object of the present invention is to insert a watermark information only for B (Bidirection) frames to suppress an increase in the amount of encoded information and to minimize deterioration of image quality. It is to provide a data copy protection device.

Another object of the present invention is to provide a moving picture data copy protection device that is inserted by different watermark information according to the position of End of Block (EOB) and mixed evenly on the original picture so that it is not easily removed by users other than the author. .

It is still another object of the present invention to provide a moving picture data copy prevention device capable of reducing hardware burden by inserting watermark information in the form of noise in a discrete cosine transformed frequency domain.

BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1 is a block diagram showing a schematic configuration of a conventional video data encoding apparatus employing a moving picture compression algorithm recommended in MPEG2.

2 is a block diagram showing a schematic configuration of a moving picture data encoding device including a moving picture data copy protection device according to the present invention;

3 is a block diagram showing an internal configuration of the digital watermark inserter of FIG.

4 is a picture structure diagram of an MPEG2 video encoding system.

* Explanation of symbols for main parts of the drawings

5: quantizer, 6: variable length coding and multiplexer,

8: inverse quantizer, 16: digital watermark inserter,

17: zigzag scanning unit, 18: switching unit,

19: address generator, 20: watermark ROM,

21: adder, 22: multiplexer,

In order to achieve the above object, the moving picture data copy protection device of the present invention selects different watermark information according to the position of the EOB representing the frequency distribution of the quantized discrete cosine transform coefficient, and watermarks only the B-type frame. A digital watermark inserter for outputting the quantized coefficients into which the information is inserted for variable length coding is provided.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to FIGS. 2 to 4 as follows.

Fig. 2 is a block diagram showing a schematic configuration of a moving picture data encoding apparatus recommended in MPEG2, and includes the moving picture data copy protection device of the present invention. In FIG. 2, blocks having the same reference numerals as those of FIG. 1 have the same configuration and operation as the above-described blocks of FIG. 1, and thus a detailed description thereof will be omitted.

A feature of the present invention is a digital watermark inserter 16 connected to the output end of the quantizer 5, wherein the output of the digital watermark inserter 16 has a variable length encoder and multiplexer 6 and an inverse quantizer ( 8) is connected. The digital watermark inserter 16 selects and inserts different watermark information according to the position of the end of block (EOB) representing the frequency distribution of the quantized discrete cosine transform coefficients. At this time, the watermark information is inserted only for the B-type frame, rather than the watermark information for all the frames of the I (Intra), P (Predictive), and B (Bidirectional) type. The quantization coefficients in which the watermark information is multiplexed are input to the variable length coding and multiplexer 6 and encoded, and also input to the inverse quantizer 8 to perform inverse quantization.

3 is a block diagram showing the internal structure of the digital watermark inserter 16. As shown in FIG. Reference numeral 17 is a zigzag scan unit, which classifies the quantization coefficients inputted from the quantizer 5 into run and level and amplitude, and corresponds to the EOB when the end of the block is reached. Detect the code to be The zigzag scanned image data is input to the switching unit 18, and the switching unit 18 is switched according to the frame type to select whether to output the input data as it is or to insert the watermark information into the input data and output the same. Done. The sign (EOB) output from the zigzag scan section 17 is input to the address generator 19, and the address generator 19 generates a different address according to the position of the input sign (EOB) to generate a watermark ROM. Output as (Watermark Rom, 20).

The watermark ROM 20 stores a plurality of digital watermark information, and reads out watermark information designated by the input address from the information, and outputs the watermark information to the adder 21. At this time, a control signal indicating a frame type is input to the terminal CS of the watermark ROM 20, and the watermark ROM 20 operates according to the control signal to read watermark information only for the B type frame. . The adder 21 connected to the output terminal of the switching unit 18 inserts the watermark information output from the watermark ROM 20 into the zigzag scanned image data and outputs the watermark information to the multiplexer 22. The multiplexer 22 multiplexes the zigzag scanned I, P type image data and the B type image data into which the watermark information is inserted in the adder 21, and then the variable length encoder and multiplexer 6 and the dequantizer ( 8).

The operation of the digital watermark inserter 16 configured as described above will be described with reference to the block diagram of FIG. 3 and the image structure diagram of FIG.

In FIG. 3, the image data input to the zigzag scan unit 17 is discrete cosine transformed by the DCT unit 4 to be converted into data in a frequency domain, and then quantized by a predetermined quantization weighting matrix in the quantizer 5. Coefficients. The quantization coefficients scanned by the zigzag scan unit 17 are classified into run and level and output to the switching unit 18. The data passed through the switching unit 18 and the multiplexer 22 is input to the variable length encoding and multiplexer 6, where encoding is performed to tie the run and the level into one code. At this time, the image data encoded in the unit of a predetermined size has no amplitude in the zigzag scanning process and continues with the run, and is encoded with a code corresponding to the EOB when the end of the block is reached.

That is, the video data encoded in every block unit ends with the set code (EOB). If the code does not end with this code (EOB), the encoding of the data should be regarded as wrong. The zigzag scan unit 17 detects a code corresponding to the EOB representing the frequency distribution of the quantized discrete cosine transform coefficient and outputs it to the address generator 19. The address generator 19 then generates an address for designating to read one of the plurality of watermark information stored in the watermark ROM 20 according to the position of the input code EOB. The watermark ROM 20 stores digital watermark information that can be inserted in a discrete cosine transformed frequency domain. For example, when encoding image data in 8 × 8 block units, 64 different watermarks are included. The information is stored.

4 is a picture structure diagram in the MPEG2 video encoding system. According to this structure diagram, an image frame is an I-type (Intraframe) that reduces spatial surplus information of the image information, a P-type (Predicted frame) that reduces the correlation between frames through Forward Prediction, and Bidirectional Prediction ) Is divided into B type (Interpolated frame) which reduces the correlation between frames. In the MPEG2 coding scheme, frames are classified into the above three types, and I, B, B, P, B, B,... , In order of I. In this case, the frames from the first I type to just before the next I type are called GOPs, and the GOP is usually composed of 12 to 15 frames.

When decoding such a moving picture, a P frame can be completely reconstructed through motion compensation only when there is a decoded preceding I frame. In addition, a B frame can be decoded through motion recovery only when there are decoded I and P frames used for prediction of the B frame. In the digital watermark inserter 16 of the present invention, watermark information is inserted only for B frames to suppress the degradation of the image quality due to the digital watermark information inserted in the form of noise and the increase in the amount of information generated during encoding. To this end, a control signal indicating a frame type is input to the watermark ROM 20 of FIG. 3 so that the watermark ROM 20 does not operate for I and P frames, and watermark information is read only for B frames.

The digital watermark information output differently according to the position of the sign EOB is input to the adder 21, and the multiplexer 22 is added with the output data of the zigzag scan unit 17 which has passed through the switching unit 18. Is sent to. At this time, the switching unit 18 is switched by the control signal indicating the frame type like the watermark ROM 20. In the case of the B frame, the output data of the adder 21 into which the watermark information is inserted is switched to the terminal a so as to be input to the multiplexer 22. In the case of I and P frames, it is switched to the terminal b so that the output data of the zigzag scan unit 17 is input to the multiplexer 22 as it is. The input data are transmitted to the variable length encoder and the multiplexer 6 and the dequantizer 8 in a multiplexed state in the multiplexer 22.

As described above, the present invention inserts only the watermark information in the form of noise into the moving picture data of the B frame in the discrete cosine transformed frequency domain, thereby reducing the burden on hardware, suppressing an increase in the amount of encoded information, and minimizing deterioration of image quality. It can be effective. In addition, since different watermark information is inserted according to the position of the EOB and mixed evenly in the frequency band of the original image, it is not easily removed by users other than the author.

Claims (5)

In the apparatus for encoding the image data of the block unit of a predetermined size through a process such as discrete cosine transform, quantization, variable length coding, motion prediction and compensation, A digital watermark that selects different watermark information according to the position of the EOB representing the frequency distribution of the quantized discrete cosine transform coefficient and outputs the variable length coding of the quantization coefficients in which the watermark information is inserted only for B-type frames. Moving picture data protection device comprising an inserter. The digital watermark inserter of claim 1, wherein A zigzag scan unit which classifies the quantized discrete cosine transform coefficients into run and level and detects a code corresponding to the EOB when the end of the block is reached; A switching unit which selects whether to output zigzag scanned image data as it is or to insert watermark information according to the frame type; An address generator for generating different addresses depending on the position of the code output from the zigzag scan unit; A watermark ROM that reads and outputs watermark information designated by an address of an address generator among a plurality of digital watermark information stored in operation according to a frame type; An adder for inserting the watermark information read from the watermark ROM into the image data after the zigzag scan is passed through the switching unit; And And a multiplexer for multiplexing the I- and P-type image data after the zigzag scan and the switching unit, and the B-type image data into which the watermark information is inserted in the adder. 3. The moving picture data according to claim 2, wherein the switching unit switches so that input data is multiplexed as it is when I and P frames are input, and is multiplexed when watermark information is inserted when a B frame is input. Copy protection. 3. The apparatus of claim 2, wherein the watermark ROM stores the number of digital watermark information that can be inserted in the discrete cosine transformed frequency domain according to the size of a block for encoding the watermark ROM. The apparatus of claim 2, wherein the watermark ROM operates according to a control signal indicating a frame type so that watermark information is read only for B frames without operating for I and P frames. .