CN101808241A - Video digital watermark embedding and detecting system based on DCT (Discrete Cosine Transformation) domain and method thereof - Google Patents

Abstract

The invention relates to a video digital watermark embedding and detecting system based on a DCT (Discrete Cosine Transformation) domain and a method thereof. The system comprises a watermark embedding device and a watermark detecting device, wherein the watermark embedding device comprises a selecting unit for selecting an embedding domain and an embedding unit for embedding legal watermark information in the embedding domain; and the watermark detecting device comprises an extracting unit for detecting and extracting the watermark information of video data to be detected and a comparing unit for identifying copyright information. By embedding the imperceptible watermark information in the digital video, the invention can detect and trace the copyright information contained in the digital video, ensures that the protected digital media content is not randomly reproduced and spread, protects digital media content suppliers and the whole industry chain, and overcomes the existing defects in the prior art.

Description

Video digital watermark embedding and detection system and method thereof based on the DCT territory

Technical field

The present invention relates to multimedia digital digital watermark field, be specifically related to a kind of video digital watermark embedding and detection system and method thereof based on the DCT territory.

Background technology

Along with popularizing of network service, many traditional media contents all change to digitlization, in wireless domain, along with the mobile network by the differentiation of the second generation to the third generation, the mobile subscriber is digital media content on the access the Internet easily and fast, and being about to based on the application of the digital media content of wired or wireless network is new tradition of information age.But the safety problem of digital media content has become bottleneck problem, is once restricting IT application process.

Ripe cryptography can solve safe transfer and access control; it is by becoming the information translation that needs protection a kind of hidden mode to stop other people to obtain; the recipient of information comes to be decoded into expressly obtaining ciphertext by corresponding key, obtains real information.

DRM (Digital Right Management, digital copyright management) is the method for another kind of copyright protection.The operation principle of DRM technology is, at first sets up the digital program authorization center, and the digital program content after the encoding compression utilizes key (Key) can encryptedly to protect (lock), and the digital program head of encryption is being deposited the URL of KeyID and program authorization center.The user according to the KeyID and the URL information of program head, just can send relevant secret key decryption (unlock) after the checking mandate by the digital program authorization center in a sowing time, and program can be play.The program that needs protection is encrypted, even preserved by user's download, the checking mandate that does not obtain the digital program authorization center also can't be play, thereby has protected the copyright of program closely.The DRM technology comprises the technical tool of digitized content intellectual property protection and management in production, propagation, sale, use, is the general name of a class technology.It is widely used in the control that conducts interviews to Digital Media.The kind of access control comprises player, broadcasting time or the time limit that limits Digital Media, forbid that recording, copy, imprinting, text duplicate etc., the multimedia kind that is suitable for comprises film, music CD, Online Music, e-book, electronic document etc., at present by Sony, apple, Microsoft, BBC's major company's extensive uses such as (BBC).DRM adds authentication information usually in multi-medium data, with the use of limited subscriber, cooperate and realize this restriction and also might add corresponding D RM module in the player, operating system, sound card.

For traditional cryptography, quick raising along with computer process ability, this method that improves system's level of confidentiality by continuous increase key length becomes more and more dangerous, in case after the deciphering, digital media content just can be random be copied, propagate, ultradistribution (Superdistribution) problem of Here it is digital media content, it has caused massive losses for media content manufacturer, thereby is restricting advancing that digital multimedia uses.

For the DRM know-why as can be seen this level of technical sophistication need common cooperation of software and hardware to finish than higher, flow process is many, control is big, and too much to user's restriction, Experience Degree is very poor.

Summary of the invention

The purpose of this invention is to provide a kind of video digital watermark embedding and detection system and method thereof based on the DCT territory; by in video data, embedding non watermark information; the video data of authorizing is carried out copyright protection; make shielded digital video content not be copied freely, propagate; complexity is lower; flow process is few, and easy to control, to overcome the deficiencies in the prior art.

For achieving the above object, the present invention adopts following technical scheme:

A kind of video digital watermark based on the DCT territory embeds and detection system, and this system comprises watermark flush mounting and watermark detection device, and described watermark flush mounting comprises: selected cell is used for the video data selection embedding territory of setting depth of decode reaching; Embed the unit, embedded digital watermarking embeds algorithm, is used for valid watermark information is embedded into described embedding territory; Described watermark detection device comprises: extraction unit, and the corresponding extraction algorithm of embedded and described embedding algorithm is used for detecting and extracting the watermark information of video data to be detected; Comparing unit, the watermark information that is used for described extraction unit is extracted is compared with described valid watermark information, and the copyright information of described video data to be detected is differentiated.

Wherein, the video data of described setting depth of decode is to be in to separate/video data of coding stage.

Wherein, described embedding territory be separate/sign indicating number compiles the quantization parameter in stage inverse discrete cosine transform/transform domain.

Wherein, described watermark detection device is integrated in FTP client FTP, is built in the Video Decoder.

Wherein, described watermark detection device is not integrated in FTP client FTP and as testing tool independently, also comprises Video Decoder.

Wherein, described embedding unit and described extraction unit are the dsp chip that embedded digital watermarking embeds algorithm and extraction algorithm.

A kind of embedding grammar of the video digital watermark based on the DCT territory, the method comprising the steps of:

S1. the video data of digital watermarking to be embedded is sent into video decoder decodes;

S2. in reaching the video data of setting depth of decode, select to embed the territory;

S3. embed algorithm according to concrete digital watermarking, valid watermark information is embedded in the described embedding territory.

Wherein, the video data of described setting depth of decode is to separate/video data of coding stage.

Wherein, described embedding territory be separate/sign indicating number compiles the quantization parameter in stage inverse discrete cosine transform/transform domain.

A kind of detection method of the video digital watermark based on the DCT territory, the method comprising the steps of:

D1. video data to be detected is decoded;

D2. detect whether comprise watermark information in the described video data, and extract described watermark information according to concrete extraction algorithm;

D3. the described watermark information that extracts is compared with valid watermark information, the copyright information of described video data to be detected is differentiated.

Beneficial effect:

1, watermark can be resisted malicious attack, uses special parameter control to embed and leaching process, after digital watermarking is embedded into, the third party who not have mandate is kept absolutely secret and can not cracks easily.

2, when watermark information exists, system has very high verification and measurement ratio, otherwise, have very low false detection rate.

3, watermarking algorithm is controlled with parameter, select the redundant information in the video to carry out the watermark embedding, and can not influence the playback rate of video, thereby farthest weaken the influence of the adding of watermark information, ensure the commercial value of former video information for the video visual quality.

4, the watermarking algorithm complexity is lower, and the real-time requirement of video data is satisfied in the embedding of digital watermarking, extraction and detection.

5, do not need original video information, directly carry out blind Detecting according to parameter.

6, watermark content can satisfy the capacity requirement of radio and television, can reach 100 bytes for real-time video, can be for video file according to file size, and maximum can reach 100KB, and watermark content can be literal or picture.

Description of drawings

Fig. 1 is that the video digital watermark that the present invention is based on the DCT territory embeds and the detection system block diagram;

Fig. 2 is the video digital watermark embedding grammar flow chart that the present invention is based on the DCT territory;

Fig. 3 for to separate/coding stage embed digital watermark schematic flow sheet;

Fig. 4 is the video digital watermark detection method flow chart that the present invention is based on the DCT territory.

Embodiment

Video digital watermark embedding and detection system and embedding and detection method based on the DCT territory that the present invention proposes are described as follows in conjunction with the accompanying drawings and embodiments.

As shown in Figure 1, the video digital watermark based on the DCT territory in the present embodiment embeds with detection system and comprises watermark flush mounting 100 and watermark detection device 200, wherein, watermark flush mounting 100 further comprises: selected cell 101 is used for the video data selection embedding territory of setting depth of decode reaching; Embedding unit 102, is the dsp chip of embedded digital watermarking embedding algorithm, is used for valid watermark information is embedded into the embedding territory; Watermark detection device 200 further comprises: extraction unit 201, and the corresponding extraction algorithm of embedded and described embedding algorithm is used for detecting and extracting the watermark information of video data to be detected; Comparing unit 202, the watermark information that is used for extraction unit 201 is extracted is compared with valid watermark information, and the copyright information of video data to be detected is differentiated.

Wherein, the video data of setting depth of decode can be original video data, or be in separate, the video data of coding stage, or the video data after the compression, embed the territory and can be original video data, or coding/decoding compile the stage discrete cosine transform (Discrete Cosine Transform, DCT)/quantization parameter in the inverse transformation territory, or the video data after the compression.Which kind of selects embed the territory, depend on the real needs to every index, this is subjected to factor restrictions such as video quality, embedding capacity, embedding speed:

Select original video data as embedding the territory, digital watermarking all can be applied to this in principle, and it is many to embed algorithm, but select original video data as embedding the territory, can increase the data bit-rate of video code flow, can lose watermark after the compression when MPEG-4 series coding, video quality can reduce.

Selection is directly embedded into watermark in the video data after the compression, the process that does not have decoding and encode again, thereby can not cause the decline of video quality, computation complexity is low simultaneously, but defines the size of watermarked data volume owing to the restriction of compression bit rate.

And select in the present embodiment watermark information is embedded into separate/quantization parameter of coding stage DCT inverse transformation/transform domain in, watermark only is embedded in the DCT coefficient, can not increase the data bit-rate of video flowing, easily designs the watermark of the multiple attack of opposing.

The video digital watermark embedding grammar based on the DCT territory that is illustrated in figure 2 as in the present embodiment comprises step:

S1. the video data of digital watermarking to be embedded is sent into the MPEG-2/MPEG-4 Video Decoder, decode according to reference format;

S2. in reaching the video data of setting depth of decode, select to embed the territory, select to be in the video data of decoding or coding stage in the present embodiment as embedding the territory;

S3. embed algorithm according to concrete digital watermarking, valid watermark information is embedded in the embedding territory, concrete embedding algorithm is the modification to the DCT DC coefficient in the present embodiment.

Because the DCT DC coefficient of the sub-piece of colourity is very sensitive to vision system, it is an existence and the very parameter of robust all the time in video flowing, this programme is embedded into watermark information the lowest order of the DC coefficient (before quantizing back, prediction) of the sub-piece DCT of colourity of video I, P and B frame after encrypting, watermark information is difficult to remove under the situation that does not influence video effect like this, so robustness is enough strong, adopt backoff algorithm to reduce mass loss simultaneously, reduce the error of bringing because of the watermark embedding.

As shown in Figure 3, the MPEG-2/MPEG-4 decoder reaches the video data decoding degree of depth when carrying out DCT inverse transformation or dct transform, and decoded data is sent to the dsp chip of built-in embedding algorithm, as stated above valid watermark information is embedded by dsp chip.

Differentiate for when producing the diffusion of mandate video data content and producing illegal copies suspicion, carrying out the copyright true and false, need the copyright information of video data is differentiated.Watermark detection device 200 in the present embodiment has two kinds of patterns:

The one, can be integrated in FTP client FTP, be built in the MPEG-2/MPEG-4 Video Decoder, when being decoded, the MPEG-2/MPEG-4 video flowing carries out the discriminating of copyright information;

Another kind of pattern is not to be integrated in the FTP client FTP, and as testing tool independently, only when the dispute over copyright and the suspicion of illegal copies occurring, use, in this case, checkout gear also comprises a MPEG-2/MPEG-4 decoder, when input MPEG-2/MPEG-4 video flowing, in the process of decoding, use watermarking algorithm that it is detected.

As shown in Figure 4, the video digital watermark detection method based on the DCT territory comprises step:

D1. MPEG-2/MPEG-4 video data to be detected is decoded;

D2. detect whether comprise watermark information in this video data, and extract the watermark information that it comprises according to concrete extraction algorithm;

D3. the watermark information that extracts is compared with valid watermark information, the copyright information of MPEG-2/MPEG-4 video data to be detected is differentiated.

The system and method that the application of the invention provides; in video data, embed non watermark information; the detection and the tracking of the copyright information that realization comprises video source; make shielded digital media content not be copied freely, propagate; protection digital media content provider and whole industry chain have solved encryption technology in " blind area " that video field is difficult to use.

Above execution mode only is used to illustrate the present invention; and be not limitation of the present invention; the those of ordinary skill in relevant technologies field; under the situation that does not break away from the spirit and scope of the present invention; can also make various variations and modification; therefore all technical schemes that are equal to also belong to category of the present invention, and scope of patent protection of the present invention should be defined by the claims.

Claims (10)

1. the video digital watermark based on the DCT territory embeds and detection system, and this system comprises watermark flush mounting (100) and watermark detection device (200), it is characterized in that,

Described watermark flush mounting (100) comprising:

Selected cell (101) is used for the video data selection embedding territory of setting depth of decode reaching;

Embed unit (102), embedded digital watermarking embeds algorithm, is used for valid watermark information is embedded into described embedding territory;

Described watermark detection device (200) comprising:

Extraction unit (201), the corresponding extraction algorithm of embedded and described embedding algorithm is used for detecting and extracting the watermark information of video data to be detected;

Comparing unit (202), the watermark information that is used for described extraction unit (201) is extracted is compared with described valid watermark information, and the copyright information of described video data to be detected is differentiated.

2. the video digital watermark based on the DCT territory as claimed in claim 1 embeds and detection system, it is characterized in that, the video data of described setting depth of decode is to be in to separate/video data of coding stage.

3. the video digital watermark based on the DCT territory as claimed in claim 1 embeds and detection system, it is characterized in that, described embedding territory be separate/sign indicating number compiles the quantization parameter in stage inverse discrete cosine transform/transform domain.

4. the video digital watermark based on the DCT territory as claimed in claim 1 embeds and detection system, it is characterized in that described watermark detection device (200) is integrated in FTP client FTP, is built in the Video Decoder.

5. the video digital watermark based on the DCT territory as claimed in claim 1 embeds and detection system, it is characterized in that, described watermark detection device (200) is not integrated in FTP client FTP and as testing tool independently, also comprises Video Decoder.

6. the video digital watermark based on the DCT territory as claimed in claim 1 embeds and detection system, it is characterized in that, described embedding unit (102) and described extraction unit (201) are the dsp chip that embedded digital watermarking embeds algorithm and extraction algorithm.

7. embedding grammar based on the video digital watermark in DCT territory is characterized in that the method comprising the steps of:

S1. the video data of digital watermarking to be embedded is sent into video decoder decodes;

S2. in reaching the video data of setting depth of decode, select to embed the territory;

S3. embed algorithm according to concrete digital watermarking, valid watermark information is embedded in the described embedding territory.

8. the embedding grammar of the video digital watermark based on the DCT territory as claimed in claim 7 is characterized in that, the video data of described setting depth of decode is to separate/video data of coding stage.

9. the embedding grammar of the video digital watermark based on the DCT territory as claimed in claim 7 is characterized in that, described embedding territory be separate/sign indicating number compiles the quantization parameter in stage inverse discrete cosine transform/transform domain.

10. detection method based on the video digital watermark in DCT territory is characterized in that the method comprising the steps of:

D1. video data to be detected is decoded;

D2. detect whether comprise watermark information in the described video data, and extract described watermark information according to concrete extraction algorithm;

D3. the described watermark information that extracts is compared with valid watermark information, the copyright information of described video data to be detected is differentiated.

Images