CN101500162B - Information embedding, extracting method and apparatus - Google Patents

Abstract

The present invention relates to a method for creating an encoding table, and the method comprises the following steps: 2<2k> index values with 2k bits are determined, the 2<2k> index values are not overlapped with each other, and k is a positive integer; 2<2k> code word groups are determined, the code word group comprises encoding code words and complement code words corresponding to the encoding code words, and the 2<2k> code word groups are not overlapped with each other; the encoding code word is provided with 2k+a bits, and a is a positive odd number; according to the index codes and the code word groups, the encoding table is created, and the encoding table comprises the one-to-one corresponding relation between the index codes and the code word groups. The present invention further relates to an encoding table creating device, an information embedding method and a device thereof, and an information extracting method and a device thereof, the methods and devices better improve the embedding efficiency of the watermark information, increase the embedding capacity, and reduce the complexity.

Description

Information embedding, extracting method and device

Technical field

The embodiment of the invention relates to technical field of information processing, relates in particular to a kind of information embedding, extracting method and device.

Background technology

Along with rapid development of multimedia; multimedia protection and authentication common attention and attention have been caused; employing information embedded technology such as digital digital watermark be to video, audio frequency, and multimedias such as image are protected and authentication is very forward position and effective method.

In information embeds; need be embedded into certain information in the carrier and go, and want to reach the purpose of protecting and authenticating, often need to embed a large amount of information; guaranteeing again simultaneously that algorithm has embeds the information invisibility preferably, does not promptly destroy the visual quality of carrier image or video.Therefore need algorithm to have higher embedding efficient, improve the amount of information that unit (as 1 bit, 1 pixel, a coefficient, a piecemeal etc.) on average can embed.

The embedding efficient that existing information embeds algorithm is often lower, especially for some special carriers, existing low embedding efficient algorithm is often not competent, for example adopt the method for matrix coder (MatrixEmbedding) at present mostly, matrix coder is that the information of carrying out of the code table according to matrix coder embeds, and it realizes that thought is as follows:

Make up code table; If there is code word a, has the bit that the n position can be revised, be used for embedding the classified information x of k bit, can construct a hash function f and from code word, extract k bit, utilize matrix coder to find code word a ' to the combination of any a and k, make x=f (a '), the Hamming distance d of a and a ' (a, a ')≤d _Max, then this code word can be expressed as (d _Max, n, k) sign indicating number, promptly the code word of a n bit changes d at most after embedding the k bit _MaxThe position; Select suitable d _Max, n, k set up the matrix coder table; Embed according to the code table information of carrying out.

Though said method has improved embedding efficient to a certain extent, the raising that embeds efficient is the cost that is reduced to the embedding capacity, and the complexity of algorithm is very high simultaneously, and the algorithm real-time is poor, and application is not strong.

Therefore lack a kind of method for building up and device and information embedding, extracting method and device that embeds efficient height, the embedding capacity is big, real-time good, complexity is low coding schedule at present.

Summary of the invention

The embodiment of the invention provides a kind of information embedding, extracting method and device, embeds efficient to improve, increases the embedding capacity, reduces complexity.

The embodiment of the invention provides a kind of information embedding method, and this method comprises:

Watermark information to be embedded is defined as index value, find the Codeword Sets that comprises coding codeword and the complement code code word corresponding with coding codeword according to index value and coding schedule, described coding schedule comprises the one-to-one relationship between described index value and the described Codeword Sets, described index value has 2k bit, described coding codeword has 2k+a bit, k is a positive integer, and a is a positive odd number; Comprise 2 in the described coding schedule ^2kIndividual index value and 2 ^2kIndividual Codeword Sets, and described 2 ^2kIndividual index value does not repeat mutually; Described 2 ^2kIndividual Codeword Sets does not repeat mutually yet;

Calculate the Hamming distance of described coding codeword and complement code code word and carrier information respectively, described carrier information is used to carry described watermark information, and the information unit number of described carrier information equals the number of bits of described coding codeword;

If the Hamming distance of described coding codeword and carrier information is less than the Hamming distance of described complement code code word and carrier information, then described coding codeword is embedded described carrier information as embedding information, otherwise the complement code code word of described coding codeword is embedded described carrier information as embedding information.

The embodiment of the invention provides a kind of information embedding device, and this device comprises:

Search module, be used for watermark information to be embedded is defined as index value, find the Codeword Sets that comprises coding codeword and the complement code code word corresponding with coding codeword according to index value and coding schedule, described coding schedule comprises the one-to-one relationship between described index value and the described Codeword Sets, described index value has 2k bit, described coding codeword has 2k+a bit, and k is a positive integer, and a is a positive odd number; Comprise 2 in the described coding schedule ^2kIndividual index value and 2 ^2kIndividual Codeword Sets, and described 2 ^2kIndividual index value does not repeat mutually; Described 2 ^2kIndividual Codeword Sets does not repeat mutually yet;

Computing module calculates the Hamming distance of described coding codeword and complement code code word and carrier information respectively, and described carrier information is used to carry described watermark information, and the information unit number of described carrier information equals the number of bits of described coding codeword;

Merge module, be used for when the Hamming distance of described coding codeword and carrier information during less than the Hamming distance of described complement code code word and carrier information, described coding codeword is embedded described carrier information as embedding information, otherwise the complement code code word of described coding codeword is embedded described carrier information as embedding information.

Above-mentioned information embedding method and device have improved the embedding efficient of watermark information preferably and embed real-time, have increased the embedding capacity, reduced complexity.

The embodiment of the invention provides a kind of information extracting method, and this method comprises:

From having the carrier information of watermark information, embedding obtains embedding information;

Find and described embedding information corresponding index value according to described embedding information and coding schedule, described index value is embedded watermark information, described embedding information comprises coding codeword or the complement code code word corresponding with coding codeword, described coding schedule comprises the one-to-one relationship between described index value and the Codeword Sets, described Codeword Sets comprises the complement code code word of coding codeword and described coding codeword, described index value has 2k bit, described coding codeword has 2k+a bit, k is a positive integer, and a is a positive odd number; Comprise 2 in the described coding schedule ^2kIndividual index value and 2 ^2kIndividual Codeword Sets, and described 2 ^2kIndividual index value does not repeat mutually; Described 2 ^2kIndividual Codeword Sets does not repeat mutually yet.

The embodiment of the invention provides a kind of information extracting device, and this device comprises:

Acquisition module is used for having the carrier information of watermark information to obtain embedding information from embedding;

The watermark extracting module, be used for finding and described embedding information corresponding index value according to described embedding information and coding schedule, described index value is embedded watermark information, described embedding information comprises coding codeword or the complement code code word corresponding with coding codeword, described coding schedule comprises the one-to-one relationship between described index value and the Codeword Sets, described Codeword Sets comprises the complement code code word of coding codeword and described coding codeword, described index value has 2k bit, described coding codeword has 2k+a bit, k is a positive integer, and a is a positive odd number; Comprise 2 in the described coding schedule ^2kIndividual index value and 2 ^2kIndividual Codeword Sets, and described 2 ^2kIndividual index value does not repeat mutually; Described 2 ^2kIndividual Codeword Sets does not repeat mutually yet.

Above-mentioned information extracting method and device can extract the watermark information of embedding quickly and efficiently from carrier information.

Below by drawings and Examples, the technical scheme of the embodiment of the invention is described in further detail.

Description of drawings

Fig. 1 is the flow chart of coding schedule method for building up embodiment of the present invention;

Fig. 2 is the structural representation of coding schedule apparatus for establishing embodiment of the present invention;

Fig. 3 is the flow chart of information embedding method embodiment one of the present invention;

Fig. 4 is the flow chart of information embedding method embodiment two of the present invention;

Fig. 5 is the structural representation of information embedding device embodiment of the present invention;

Fig. 6 is the flow chart of information extracting method embodiment of the present invention;

Fig. 7 is the structural representation of information extracting device embodiment of the present invention.

Embodiment

As shown in Figure 1, be the flow chart of coding schedule method for building up embodiment of the present invention, this method comprises:

Step 101, determine 2 ^2kIndividual index value with 2k bit, above-mentioned 2 ^2kIndividual index value does not repeat mutually, and k is a positive integer;

Always there be a kind of in the following situation for the long bit sequence of any one odd number:

1. complete 0; 2. complete 1; 3. 0 number is more than 1; 4. 1 number is more than 0

For example: for bit sequence

From right to left, the position at each bit place is called P ₁(position 1), P ₂(position 2) ... P _2K+1(position 2k+1), when the number when 0 was less than 1 number, 0 was called as minority, and the position at 0 place is the minority position, on the contrary 1 is called as minority, and the position at 1 place is the minority position; Be 0 entirely or be that 1 situation is special circumstances entirely.For example: length is 5 bit sequence: ' 11010 ', the value at position 1 and 3 places, position is 0, the value at position 2,4 and 5 places be 1,1 number more than 0,0th, minority wherein can definition position 1 and 3 be " minority position " like this; Same 11010 complementary series (step-by-step negate) ' 00101 ', the value at position 1 and 3 places, position is 1, the value at position 2,4 and 5 places is 0, then 0 number is more than 1, the 1st, minority wherein, can definition position 1 and 3 be " minority position " equally, the minority position of visible complementary series be the same.

Hence one can see that, for given bit sequence, any combination of " minority position " corresponding two kinds of values of this bit sequence, for example, it to length 5 bit sequence, this combined situation of P1 and P3 can exist two corresponding complementary seriess (11010 and 00101) to comprise the combined situation (being that P1P3 belongs to the minority position) of this minority as described above; When the bit sequence is complete 0 or complete 1 the time, " the minority position " of bit sequence do not exist, and " minority position " number is 0, can call this situation the special circumstances of " minority position ";

According to the knowledge of permutation and combination, the number of all possible minority position (comprising special circumstances) is in a long bit sequence of odd number:

$C_{2k+1}^0+C_{2k+1}^1+...+C_{2k+1}^k=\underset{i=0}{\overset{k}{\mathrm{\&Sigma;}}}{C}_{2k+1}^i=2^{2k}...\left(1\right)$

The value of k can be any positive integer herein, supposes that k is 1 herein, and the index value that then has 2 bit positions has 00,01,10,11 4 kind of form;

Step 102, determine 2 ^2kIndividual Codeword Sets, above-mentioned Codeword Sets comprise coding codeword and the complement code code word corresponding with above-mentioned coding codeword, above-mentioned 2 ^2kIndividual Codeword Sets does not repeat mutually; Above-mentioned coding codeword has 2k+a bit, and a is a positive odd number;

Above-mentioned k can be any positive integer, a can be any positive odd number, satisfy relation one to one as long as comprise the Codeword Sets of the coding codeword of 2k+a bit with the index value with 2k bit, the embodiment of the invention all is that example describes with a=1 for convenience of description.Then be example with k=1,2k+1=3, code word length is 3, according to formula (1), the combination of " minority position " has 4 kinds of situations (having comprised special circumstances), and the bit sequence (code word) of every kind of corresponding two complementations of situation is designated as these 4 kinds of situations respectively: 0,1,2,3 as index value.Specifically as shown in table 1, code word and the complement code code word thereof with 3 information bits can be divided into four groups, for example 000 and 111 1 group, 011 and 100 1 group, 101 and 010 1 groups, 110 and 001 1 groups.

Minority's position grouping situation during table 1k=1

Minority's position number	Minority's position grouping	Index value	Codeword Sets
				0	Do not exist	0	000 or 111
1	P1	1	011 or 100

?	P2	2	101 or 010
				?	P3	3	110 or 001

During with k=2, sequence length is 5, and the number of minority position can be 1, also can be 2, and according to formula (1), totally 16 kinds of combined situation are designated as an index to each situation, are respectively: 0,1,2 ..., 15, specifically as shown in table 2.

Minority's position grouping situation during table 2k=2

Minority's position number	Minority's position grouping	Index value	Codeword Sets
				0	Do not exist	0	00000 or 11111
1	P1	1	10000 or 01111
				1	P2	2	01000 or 10111
1	P3	3	00100 or 11011
				1	P4	4	00010 or 11101
1	P5	5	00001 or 11110
				2	P5，P4	6	11000 or 00111
2	P5，P3	7	10100 or 01011
				2	P5，P2	8	10010 or 01101
2	P5，P1	9	10001 or 01110
				2	P4，P3	10	01100 or 10011
2	P4，P2	11	01010 or 10101
				2	P4，P1	12	01001 or 10110
2	P3，P2	13	00110 or 11001
				2	P3，P1	14	00101 or 11010
2	P2，P1	15	00011 or 11100

Step 103, set up coding schedule according to above-mentioned index value and Codeword Sets, above-mentioned coding schedule comprises the one-to-one relationship between above-mentioned index value and the above-mentioned Codeword Sets.

With k=1 is example, can set up corresponding relation to above-mentioned four index values and four Codeword Sets in any order, for example as shown in table 3, the coding schedule of setting up during for present embodiment k=1; Certainly, index value and Codeword Sets are as long as satisfy relation one to one, and concrete corresponding content can be adjusted as required.

Coding schedule during table 3k=1

Index	Codeword Sets
		00	000 or 111
01	011 or 100
		10	101 or 010
11	110 or 100

In like manner, it is as shown in table 4 below during k=2, can to obtain " minority's coding " code table equally.

Coding schedule during table 4k=2

Index	Codeword Sets
		0000	00000 or 11111
0001	10000 or 01111
		0010	01000 or 10111
0011	00100 or 11011
		0100	00010 or 11101
0101	00001 or 11110
		0110	11000 or 00111
0111	10100 or 01011
		1000	10010 or 01101
1001	10001 or 01110
		1010	01100 or 10011
1011	01010 or 10101

1100	01001 or 10110
		1101	00110 or 11001
1110	00101 or 11010
		1111	00011 or 11100

Above-mentioned coding schedule method for building up embodiment, all exist for the long bit sequence of any one odd number be 0 entirely, be 1,0 number entirely than 1 lack, 1 number is than 0 situation about lacking, except being 0 entirely, being 1 special circumstances entirely, in two code words of above-mentioned any one Codeword Sets, all have 0 number than 1 lack, 1 number lacks two kinds of situations than 0, above-mentioned situation is laid a good foundation for the embedding efficient that guarantees embedding information.For example, code word during coding has two kinds of selections, when watermarked, the code word that can select carrier information to be made less modification is carried out the watermark embedding like this, thereby improves the embedding efficient (embedding information bit and the ratio that is modified the carrier information bit number) of watermarking algorithm.

As shown in Figure 2, be the structural representation of coding schedule apparatus for establishing embodiment of the present invention, this device comprises: index value unit 11 is used for determining 2 ^2kIndividual index value with 2k bit, above-mentioned 2 ^2kIndividual index value does not repeat mutually, and k is a positive integer; Codeword Sets unit 12 is used for determining 2 ^2kIndividual Codeword Sets, above-mentioned Codeword Sets comprise the complement code code word of coding codeword and above-mentioned coding codeword correspondence, above-mentioned 2 ^2kIndividual Codeword Sets does not repeat mutually; Above-mentioned coding codeword has 2k+a bit, and a is a positive odd number; Code table is set up unit 13, is used for setting up coding schedule according to above-mentioned index value and Codeword Sets, and above-mentioned coding schedule comprises the one-to-one relationship between above-mentioned index value and the above-mentioned Codeword Sets.

Above-mentioned a can be any positive odd number, satisfies relation one to one with the index value with 2k bit, preferred a=1 herein as long as comprise the Codeword Sets of the coding codeword of 2k+a bit.

Above-mentioned coding schedule apparatus for establishing embodiment has set up one-to-one relationship with index value and Codeword Sets, for the embedding efficient and the extraction efficiency of the information of embedding that guarantee information to be embedded are laid a good foundation.

Embodiment of the invention information embedding method comprises:

Watermark information to be embedded is defined as index value, find the Codeword Sets that comprises coding codeword and the complement code code word corresponding with coding codeword according to index value and coding schedule, above-mentioned coding schedule comprises the one-to-one relationship between above-mentioned index value and the above-mentioned Codeword Sets, above-mentioned index value has 2k bit, above-mentioned coding codeword has 2k+a bit, k is a positive integer, and a is a positive odd number;

Calculate the Hamming distance of above-mentioned coding codeword and complement code code word and carrier information respectively, above-mentioned carrier information is used to carry above-mentioned watermark information, and the information unit number of above-mentioned carrier information equals the number of bits of above-mentioned coding codeword;

If the Hamming distance of above-mentioned coding codeword and carrier information is less than the Hamming distance of above-mentioned complement code code word and carrier information, then above-mentioned coding codeword is embedded above-mentioned carrier information as embedding information, otherwise the complement code code word of above-mentioned coding codeword is embedded above-mentioned carrier information as embedding information.

Simultaneously, preferred in the present embodiment a=1.

As shown in Figure 3, be the flow chart of information embedding method embodiment one of the present invention, this method comprises:

Step 201, carrier information is divided into groups by every group of 2k+1 unit, watermark information to be embedded is divided into groups by every group of 2k bit;

Above-mentioned carrier information can be image information or video information etc., be assumed to be image information herein, the set of pixel that then will be to be revised is designated as X, the watermark bit sequence that embeds is designated as W, X is divided into one group of every 2k+a pixel, and a is positive odd number, a=1 herein, W is divided into one group of every 2k bit, and the value of above-mentioned k can be positive integer;

Step 202, according to the extracting method of watermark bit in the watermarking algorithm, one of random extraction is the information bit in the carrier information grouping of embed watermark information;

Can be according to the extracting method of watermark bit in any watermarking algorithm such as digital picture watermarking algorithm, Audio Watermarking Algorithm, the video watermark algorithm, extract in certain grouping the information bit in 2k+1 the pixel, the information bit in this 2k+1 unit is designated as A, A={A _i(i=[1,2k+1]), the value of supposing k is 1, and the watermarking algorithm of herein selecting is supposed to obtain A=111 according to the watermark bit extracting method of LSB watermarking algorithm for replacing least significant bit (LSB) algorithm;

Step 203, select 2k bit in the grouping of watermark information to be embedded as index value at random, find the Codeword Sets that comprises coding codeword and the complement code code word corresponding according to coding schedule with coding codeword;

Above-mentioned coding schedule is the index value set up according to the operating procedure of the method for building up embodiment of a coding schedule of the present invention and Codeword Sets coding schedule one to one, and above-mentioned coding codeword and the complement code code word corresponding with this coding codeword are designated as B and C respectively; Suppose W certain the grouping in bit be 00, according to table 1, index 00 can be encoded as B=000 or C=111;

Hamming distance D1 between step 204, comparison A and the B and the size of the Hamming distance D2 between A and the C, less than D2, then execution in step 205 as if D1, otherwise execution in step 206;

Step 205, watermark information to be embedded are encoded to B,

Step 206, watermark information to be embedded are encoded to C;

Hamming distance D1=3 between A and the B, the Hamming distance D2=0 between A and the C is because D2 is less than D1, so watermark information to be embedded is encoded to C;

Step 207, replace the carrier information grouping of above-mentioned random extraction with the watermark information behind the coding;

C=111 is embedded among the A=111 of said extracted, but does not need to revise any pixel,, then need to revise 3 pixels if B is embedded among the A=111 of said extracted;

Step 208, judge that whether all watermarks grouping all embeds in the carrier information, if then finish, otherwise turns to step 202.

Repeating aforesaid operations all adopts the mode of revising the least possible pixel to be embedded in the carrier information grouping until all watermark groupings.

Said method, utilize two code words of each Codeword Sets and the original image information position of extraction to compare, select to revise the few code word of pixel and replace the original image information position of extracting, improved the embedding efficient of watermark information preferably and embed real-time, increased the embedding capacity, reduced complexity.

With k=2 is example, and then index value is 4 bits, and coding codeword is 5 bits.Suppose that watermark information is divided into 0101,0010,1,010 three groups, carrier information information promptly to be revised divides 01001,00110,10,011 3 groups.As index value, the coding schedule of question blank 4 obtains decode word group and is respectively 00001 or 11110,01000 or 10111,01100 or 10011 with three groups of watermark informations; The Hamming distance of each coding codeword of compute codeword group and complement code and carrier information corresponds to 1 or 4,3 or 2,5 or 0 respectively; Because 1＜4,3＞2,5＞0, thus select respectively with Hamming distance 1,2,0 corresponding codes code word or complement code code word as embedding information, be respectively 00001,10111,10011.Revise carrier information according to embedding information, in the present embodiment carrier information is revised as embedding information.Make that like this bit that needs to revise only is the bit of the position of aforementioned minority, has improved embedding efficient and embedding capacity greatly.

With the example that is applied as in video watermark process, as shown in Figure 4, be the flow chart of information embedding method embodiment two of the present invention, this method comprises:

Step 301, parsing code stream, the motion vector of the some estimating motion pieces in each frame of selection video sequence is as watermark host's vector to be embedded;

Step 302, on the horizontal component of each motion vector or vertical component embed watermark information;

On above-mentioned horizontal component or vertical component, embed 1 bit watermark information;

Step 303, determine to comprise the judgement formula V of motion vector component, the judgement formula is carried out Modulo-two operation, its result is 0 or 1;

Step 304, judge whether to obtain all results, if execution in step 305 then, otherwise turn to step 301;

Repeat above-mentioned steps 301-303, until obtaining all results;

Step 305, all results' set is designated as F;

Step 306, F and watermark information W are divided into groups;

Select k=1, F is divided into one group of per 3 bit, watermark information W is divided into one group of per 2 bit;

Step 307, according to coding schedule, based on watermark information W, F is revised as F ';

According to coding schedule table 3,, revise former set F based on watermark information W to be embedded;

Step 308, the frame of selecting needs to revise according to the difference of F and F ' embed in this frame of rules modification a certain motion vector according to former algorithm, make the result of amended judgement formula V mould 2 satisfy F ', embed thereby finish watermark information;

Step 309, judge whether to embed all watermark informations, if then finish, otherwise, turn to step 307;

Repeat above-mentioned steps 307-308, until embedding all watermark informations.

With the example that is applied as in the digital figure watermark algorithm, for example the information embedding method that the embodiment of the invention is provided is added and subtracted 1 algorithm at random with LSB and is combined, and can effectively improve the quality that embeds efficient and watermarked back image.It is as follows that the watermark information that LSB adds and subtracts 1 algorithm at random embeds rule: if the watermark information bit is identical with this pixel lowest bit position during embedding, then this pixel value remains unchanged; If the watermark information bit is different with this pixel lowest bit position, then this pixel value is added at random 1 or subtract 1.Above-mentioned watermark information is the watermark information of determining according to minority's coding code table through being, table 5 show adopt LSB at random addition and subtraction cooperate minority's coding method and the former LSB embedding efficient situation of addition and subtraction at random, adopted standard picture lena, baboon, peppers, airplane as carrier image, the image size is 256 * 256, and embed watermark information is that size is 400 * 70 bianry image.

Table 5 minority coding method experimental result situation

As can be seen from Table 5, after the use " minority's coding ", the embedding efficient of watermark information obviously improves, and mostly improves more than 33% on the original basis.Said method, make full use of the watermark information coding method of coding schedule method for building up embodiment structure of the present invention, select to cause the code word of less modification according to the situation of host's medium and watermarking algorithm, thereby improved the embedding efficient of watermark information preferably and embed real-time, increased the embedding capacity, reduced complexity.

As shown in Figure 5, structural representation for information embedding device embodiment of the present invention, this device comprises: search module 21, be used for watermark information to be embedded is defined as index value, find the Codeword Sets that comprises coding codeword and the complement code code word corresponding with coding codeword according to index value and coding schedule, above-mentioned coding schedule comprises the one-to-one relationship between above-mentioned index value and the above-mentioned Codeword Sets, above-mentioned index value has 2k bit, above-mentioned coding codeword has 2k+a bit, k is a positive integer, and a is a positive odd number; Computing module 22 calculates the Hamming distance of above-mentioned coding codeword and complement code code word and carrier information respectively, and above-mentioned carrier information is used to carry above-mentioned watermark information, and the information unit number of above-mentioned carrier information equals the number of bits of above-mentioned coding codeword; Merge module 23, be used for when the Hamming distance of above-mentioned coding codeword and carrier information during less than the Hamming distance of above-mentioned complement code code word and carrier information, above-mentioned coding codeword is embedded above-mentioned carrier information as embedding information, otherwise the complement code code word of above-mentioned coding codeword is embedded above-mentioned carrier information as embedding information.

Wherein, the preferred value of a is 1.

Above-mentioned information embedding device, utilization is searched module and is determined index value, calculate the Hamming distance of above-mentioned coding codeword and complement code code word and carrier information more respectively by computing module, by merge module above-mentioned coding codeword is embedded above-mentioned carrier information as embedding information at last, thereby finish the telescopiny of watermark information, said apparatus has improved the embedding efficient of watermark information preferably and has embedded real-time, has increased the embedding capacity, reduced complexity.

Embodiment of the invention information extracting method comprises:

From having the carrier information of watermark information, embedding obtains embedding information;

Find and above-mentioned embedding information corresponding index value according to above-mentioned embedding information and coding schedule, above-mentioned index value is embedded watermark information, above-mentioned embedding information comprises coding codeword or the complement code code word corresponding with coding codeword, above-mentioned coding schedule comprises the one-to-one relationship between above-mentioned index value and the Codeword Sets, above-mentioned Codeword Sets comprises the complement code code word of coding codeword and above-mentioned coding codeword, above-mentioned index value has 2k bit, above-mentioned coding codeword has 2k+a bit, k is a positive integer, and a is a positive odd number.

Simultaneously, preferred in the present embodiment a=1.

As shown in Figure 6, be the flow chart of information extracting method embodiment of the present invention, this method comprises:

Step 401, the value of k when embedding according to watermark information are divided into groups the carrier information behind the embed watermark information by every group of 2k+1 unit;

Carrier information behind the acquisition embed watermark information, above-mentioned carrier information can be for image, audio frequency, video etc., when carrier information is image, the set of amended pixel in the carrier information is designated as X ', k value when embedding according to watermark is divided into one group of every 2k+1 pixel to X ';

Information bit in the carrier information grouping of step 402, an embed watermark information of random extraction is as coding codeword, according to Codeword Sets and index value one to one coding schedule find and the Codeword Sets corresponding index value that comprises above-mentioned coding codeword, above-mentioned index value is a grouping of embedded watermark information;

Extract the information bit of 2k+1 pixel in certain X ' grouping, be designated as P={Pi} (i=[1,2k+1]); The code word of P as coding schedule, the coding schedule of setting up during according to embed watermark information finds and the Codeword Sets corresponding index value that comprises above-mentioned code word, and then index value is a grouping in the embedded watermark information;

Step 403, whether extract all watermark informations, if then finish, otherwise, turn to step 401.

Repeat aforesaid operations until from carrier information, extracting all embeddings watermark information wherein.

Above-mentioned information extracting method, the coding schedule that has adopted when having utilized information to embed preferably can extract the watermark information of embedding quickly and efficiently from carrier information according to the one-to-one relationship of Codeword Sets in the coding schedule and index value.

As shown in Figure 7, be the structural representation of information extracting device embodiment of the present invention, this device comprises: acquisition module 31 is used for having the carrier information of watermark information to obtain embedding information from embedding; Watermark extracting module 32, be used for finding and above-mentioned embedding information corresponding index value according to above-mentioned embedding information and coding schedule, above-mentioned index value is embedded watermark information, above-mentioned embedding information comprises coding codeword or the complement code code word corresponding with coding codeword, above-mentioned coding schedule comprises the one-to-one relationship between above-mentioned index value and the Codeword Sets, above-mentioned Codeword Sets comprises the complement code code word of coding codeword and above-mentioned coding codeword, above-mentioned index value has 2k bit, above-mentioned coding codeword has 2k+a bit, k is a positive integer, and a is a positive odd number.

Wherein, the preferred value of a is 1.

Above-mentioned information extracting device, the coding schedule that has adopted when having utilized information to embed preferably can extract the watermark information of embedding quickly and efficiently from carrier information according to the one-to-one relationship of Codeword Sets in the coding schedule and index value.

It should be noted that at last: above embodiment only in order to technical scheme of the present invention to be described, is not intended to limit; Although with reference to previous embodiment the present invention is had been described in detail, those of ordinary skill in the art is to be understood that: it still can be made amendment to the technical scheme that aforementioned each embodiment put down in writing, and perhaps part technical characterictic wherein is equal to replacement; And these modifications or replacement do not make the essence of appropriate technical solution break away from the spirit and scope of various embodiments of the present invention technical scheme.

Claims (8)

1. information embedding method is characterized in that comprising:

Watermark information to be embedded is defined as index value, find the Codeword Sets that comprises coding codeword and the complement code code word corresponding with coding codeword according to index value and coding schedule, described coding schedule comprises the one-to-one relationship between described index value and the described Codeword Sets, described index value has 2k bit, described coding codeword has 2k+a bit, k is a positive integer, and a is a positive odd number; Comprise 2 in the described coding schedule ^2kIndividual index value and 2 ^2kIndividual Codeword Sets, and described 2 ^2kIndividual index value does not repeat mutually; Described 2 ^2kIndividual Codeword Sets does not repeat mutually yet;

Calculate the Hamming distance of described coding codeword and complement code code word and carrier information respectively, described carrier information is used to carry described watermark information, and the information unit number of described carrier information equals the number of bits of described coding codeword;

If the Hamming distance of described coding codeword and carrier information is less than the Hamming distance of described complement code code word and carrier information, then described coding codeword is embedded described carrier information as embedding information, otherwise the complement code code word of described coding codeword is embedded described carrier information as embedding information.

2. information embedding method according to claim 1 is characterized in that described a equals 1.

3. information embedding device is characterized in that comprising:

Search module, be used for watermark information to be embedded is defined as index value, find the Codeword Sets that comprises coding codeword and the complement code code word corresponding with coding codeword according to index value and coding schedule, described coding schedule comprises the one-to-one relationship between described index value and the described Codeword Sets, described index value has 2k bit, described coding codeword has 2k+a bit, and k is a positive integer, and a is a positive odd number; Comprise 2 in the described coding schedule ^2kIndividual index value and 2 ^2kIndividual Codeword Sets, and described 2 ^2kIndividual index value does not repeat mutually; Described 2 ^2kIndividual Codeword Sets does not repeat mutually yet;

Computing module calculates the Hamming distance of described coding codeword and complement code code word and carrier information respectively, and described carrier information is used to carry described watermark information, and the information unit number of described carrier information equals the number of bits of described coding codeword;

Merge module, be used for when the Hamming distance of described coding codeword and carrier information during less than the Hamming distance of described complement code code word and carrier information, described coding codeword is embedded described carrier information as embedding information, otherwise the complement code code word of described coding codeword is embedded described carrier information as embedding information.

4. information embedding device according to claim 3 is characterized in that described a equals 1.

5. information extracting method is characterized in that comprising:

From having the carrier information of watermark information, embedding obtains embedding information;

Find and described embedding information corresponding index value according to described embedding information and coding schedule, described index value is embedded watermark information, described embedding information comprises coding codeword or the complement code code word corresponding with coding codeword, described coding schedule comprises the one-to-one relationship between described index value and the Codeword Sets, described Codeword Sets comprises the complement code code word of coding codeword and described coding codeword, described index value has 2k bit, described coding codeword has 2k+a bit, k is a positive integer, and a is a positive odd number; Comprise 2 in the described coding schedule ^2kIndividual index value and 2 ^2kIndividual Codeword Sets, and described 2 ^2kIndividual index value does not repeat mutually; Described 2 ^2kIndividual Codeword Sets does not repeat mutually yet.

6. information extracting method according to claim 5 is characterized in that described a equals 1.

7. information extracting device is characterized in that comprising:

Acquisition module is used for having the carrier information of watermark information to obtain embedding information from embedding;

The watermark extracting module, be used for finding and described embedding information corresponding index value according to described embedding information and coding schedule, described index value is embedded watermark information, described embedding information comprises coding codeword or the complement code code word corresponding with coding codeword, described coding schedule comprises the one-to-one relationship between described index value and the Codeword Sets, described Codeword Sets comprises the complement code code word of coding codeword and described coding codeword, described index value has 2k bit, described coding codeword has 2k+a bit, k is a positive integer, and a is a positive odd number; Comprise 2 in the described coding schedule ^2kIndividual index value and 2 ^2kIndividual Codeword Sets, and described 2 ^2kIndividual index value does not repeat mutually; Described 2 ^2kIndividual Codeword Sets does not repeat mutually yet.

8. information extracting device according to claim 7 is characterized in that described a equals 1.

Images