CN104484587A - Mobile phone photographing authentication method - Google Patents

Abstract

The invention provides a mobile phone photographing authentication method, and belongs to the technical field of digital image processing. The mobile phone photographing authentication method provided by the invention comprises a signature process and an authentication process. The signature process realizes the signature function of embedding authentication information such as watermark in the image. The signature process mainly comprises the implementation processes of chaotic watermark generation, mobile phone signal generation, embedded watermark generation, image pre-treatment, watermark embedding, signature image generation and the like. The authentication process realizes the process for verifying whether the authentication image is manipulated maliciously (the relatively serious malicious manipulation can be detected, and non-malicious slight correction can pass), and verifying the functions of copyright authentication, information source validity and the like. The authentication process is the inverse process of the signature process, the process realizes the extraction of watermarks and the phone number information extracted according to the watermark is compared with the initial phone number, so as to realize the specific function of image authentication.

Description

A kind of method of mobile phone photograph certification

Technical field

This patent belongs to digital image processing techniques category.Relate generally to the correlative technology fields such as mobile computing, digital copyright protecting, pattern-recognition, information security.

Background technology

The rise of smart mobile phone is the life of people, study, work provide great convenience, and people can use smart mobile phone to process daily simple affairs very easily.Mobile phone is integrated with the function of routine office work equipment more and more, and current smart mobile phone instead of computer, camera, scanner in certain occasion, for people save a large amount of time, energy, financial cost.But along with exchanging more and more easily conveniently, the image information of mobile phone terminal collection is also more and more easily tampered, encroaches right.At present, the mode great majority for the process of image concentrate on computer end or network canbe used on line.

Digital watermark technology is mainly adopted to realize in computer end.

Summary of the invention

Present invention achieves a kind of method for mobile phone photograph certification.

The method of the mobile phone photograph certification that patent of the present invention proposes, comprises signature process and verification process two steps.

Signature process realizes the signature function being embedded in the authentication informations such as watermark at image.Signature process mainly comprises: the implementation processes such as chaotic watermarking generation, the generation of cell-phone number signal, embed watermark generation, Image semantic classification, watermark embedment, signature image generation.Verification process realizes the proof procedure of the functions such as whether authentication image is maliciously tampered (malice that degree is more serious is distorted and can be detected, the slight modifications of non-malicious can by checking), copyright authentication, information source validity.Verification process is the extraction of the inverse process to signature process, this process implementation watermark, to according to this watermark extracting cell-phone number information out and the comparison of original cell-phone number information, thus realizes the concrete function of image authentication.

Accompanying drawing explanation

Figure embeds some schematic diagram, and the point of P shown in it is embed watermark point.

Embodiment

Project implementation specifically comprises signature process and verification process two steps.

Signature process mainly comprises: the implementation processes such as chaotic watermarking generation, the generation of cell-phone number signal, embed watermark generation, Image semantic classification, watermark embedment, signature image generation.

Verification process is the extraction of the inverse process to signature process, this process implementation watermark, to according to this watermark extracting cell-phone number information out and the comparison of original cell-phone number information, thus realizes the concrete function of image authentication.

The performing step of signature process mainly contains:

1. chaos Computer image genration

The privately owned value that this process sets according to a user generates a width two-value chaos image as secret key.This process use chaos sequence is,

$\begin{array}{cc}{x}_{n+1}=1-2x_n^2& x_n\∈[-\mathrm{1,1}]\end{array}$

Use initial value is x ₀, this value is the secret key that user sets.In formula, n indicates the length of production sequence, n=M*N+a.

Above-mentioned formula generates a chaos sequence, intercepts from a+1 data above-mentioned sequence.Binaryzation is carried out to the sequence intercepted, is specially,

$x_n=\left\{\begin{array}{cc}1& x_n>=0\\ 0& \mathrm{else}\end{array}\right.$

Above-mentioned formula generates a binary sequence, is the image array CW of M*N by its transposition.

2. cellphone information Computer image genration

Cell-phone number is treated to

x ₀＝Number*10 ^-13

By the x obtained ₀the image array NW of a M*N size is generated as the initial value in step 1.

3. embed watermark generates

The image array C that step 1 is generated ₀the image array N generated with step 2 ₀carry out XOR process, use formula is,

W＝XOR(CW,NW)

In formula, XOR is xor operation, and W is the watermarking images information for embedding generated.

4. Image semantic classification

Image semantic classification mainly carries out wavelet transformation to image, selects the particular sub-band of wavelet transformation, in subband, carries out Blind extracting average value processing, uses the processes such as chaos selected embedding point.

4.1 wavelet transformation

To needing the image I signed to carry out lifting wavelet transform, obtaining four subbands, being specially:

(W _LL,W _HL,W _LH,W _HH)＝LFT(I _i,j)

In formula, LL is low frequency sub-band, and low frequency carries the significant energy having concentrated image.

4.2 replace center pixel value

Divide low frequency sub-band LL, be divided into the sub-block of 8*8 size, using the central point of this sub-block as embedding point, P point is embed watermark point as shown in the figure.Calculate the pixel value of center point P in this sub-block respectively, this pixel value is replaced with the average of eight pixels around it, specific implementation is:

${\mathrm{Value}}_P=\frac{1}{8}\underset{i=0}{\overset{7}{\mathrm{\Σ}}}{\mathrm{Value}}_i$

4.3 determine to embed pixel

According to the Logistic formula in step 1, determine that a sequence length is the sequence of present image size size (I)/8 pixel size, it is sorted, then get the sequence postorder position be listed in its original sequence and form a new sequence N2.Intercept from this sequence the n-th 1 pixels, intercepting its length is M*N, forms new sequence N3.

Central pixel point all in step 4.2 be numbered, the sequence number of being numbered in N3 is defined as embed watermark point.

5 embed watermarks

Embed watermark information on the position determined, embedded mode is,

E＝Value _p,ij+α*W _i,j

In formula, E is the pixel value after embed watermark, Value _{p, ij}really certain pixel on allocation, W _i,jthe watermark information that to be position be on (i, j), α is watermark embed strength.

After above-mentioned steps completes, in low frequency sub-band LL, complete embedding, remember that newly-generated low frequency sub-band is W _lLN.

6. generate signature image

For the LLN generated in step 5, carry out inverse lifting wavelet transform, formula is,

IN＝ILFT(W _LLN,W _HL,W _LH,W _HH)

In formula, IN is the image after signature, and ILFT () is inverse lifting wavelet transform, and WLLN is the low frequency sub-band of embed watermark information in step 5, W _hL, W _lH, W _hHit is the subband that in step 4.1, lifting wavelet transform obtains.

Verification process is responsible for extracting watermark information in signature image, and has compared the function of certification to watermark information.It specifically implements to mainly contain following steps:

1. lifting wavelet transform

This process uses identical formula manipulation with the step 4.1 of signature process.Low frequency wavelet subband can be obtained after having processed, be labeled as LLC.

2. extract watermark information

According to the mode determination embedded location introduced in signature process step 4, extract watermark information from these positions.

Determining position, first calculating the average that this pixel closes on eight neighborhood, then deducting this average with current pixel value, after intensity process, obtain watermark information, be specially,

WC _i,j＝(E-ValueC _p,ij)/α

In formula, WC _i,jthe watermark information extracted, ValueC _{p, ij}be eight neighborhood average, E is current pixel value, and α is watermark embed strength.

3. generate chaos image

Chaos image CWC is generated according to the mode of step 1 in signature process.

4. watermarking images extracts

The WC obtained in applying step 2, the CWC that step 3 obtains carries out xor operation, obtains the watermark information WI proposing out.Use formula to be,

WI＝XOR(CWC,WC)

5. certification

Original cellphone information image NW when the watermarking images WI extract verification process and signature compares, and determines the authentication result of present image according to comparative result.Be specially,

In formula, T is the threshold value of setting, and this threshold value is set by authenticating party.

Claims (2)

1. a method for mobile phone photograph certification, is characterized in that his processing procedure is:

Project implementation specifically comprises signature process and verification process two steps; Signature process mainly comprises: the implementation processes such as chaotic watermarking generation, the generation of cell-phone number signal, embed watermark generation, Image semantic classification, watermark embedment, signature image generation;

Key of Implementation step is:

1) cellphone information Computer image genration

Cell-phone number is treated to

x ₀＝Number*10 ^-13

By the x obtained ₀the image array NW of a M*N size is generated as the initial value in step 1;

2) embed watermark generates

The image array that image array and watermark generate is carried out XOR process, the watermarking images information for embedding of generation;

3) center pixel value is replaced

Low frequency sub-band is divided, is divided into the sub-block of 8*8 size, using the central point of this sub-block as embedding point embed watermark information.Calculate the pixel value of central point in this sub-block respectively, this pixel value is replaced with the average of eight pixels around it, specific implementation is:

${\mathrm{Value}}_P=\frac{1}{8}\underset{i=0}{\overset{7}{\mathrm{\Σ}}}{\mathrm{Value}}_i$

2. the method for a kind of mobile phone photograph certification according to claim 1, is characterized in that: the original cellphone information image NW when watermarking images WI of extraction and signature compares, and is expressed as

In formula, T is the threshold value of setting, and this threshold value is set by authenticating party.