IRJET- A Study on Image Authentication Methods
0 likes19 views
This document discusses and compares different methods for authenticating digital images, including cryptographic authentication, robust image hashing authentication, and watermarking authentication. Cryptographic authentication uses encryption and decryption but has limitations in localization. Robust image hashing is based on rotation-invariant moments but struggles with large area cropping. Watermarking authentication allows for embedding watermarks for identification and tamper detection but cannot precisely locate tampered regions. Overall, the document concludes that watermarking schemes constructed using chaotic maps provide better security compared to other methods due to chaotic maps' sensitivity to initial conditions.
![International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 05 Issue: 12 | Dec 2018 www.irjet.net p-ISSN: 2395-0072
© 2018, IRJET | Impact Factor value: 7.211 | ISO 9001:2008 Certified Journal | Page 1720
3.2 Robust image hashing authentication
Image hashing is used to identify the duplicate copies of the
original images. A robust hashing technique is used for
detecting imageforgery,colourmodification,andforlocating
the forged area. The robust hashing method is based on
Random Transform performed on input image to obtain the
projections in various orientations. The insignificant
coefficients are removed. The invariant moments are
calculated. DFT is performed on the invariant moments.
Finally, the image hash value H is defined as the normalized
and quantized version of the significant DFT coefficientsand
then the hash bits are generated. Most of the images hashing
algorithms have their disadvantages in getting the desirable
performance against a particular image processing attack.
The disadvantage of robust image hashing is large area
cropping.
3.3 Watermarking authentication
Digital watermarking involves embedding information into
digital multimedia content such that the information can
later can be extracted or detected by the recipient for
different purposes such as control and copy prevention.The
process of watermarking has to be resilient against
tampering attacks, keeping the content of a watermark
readable in order to be recognizable when extracted by the
recipient. The significant features of watermarking system
are robustness and fidelity. The size of the embedded
information has to be considered. The data becomes less
robust as its size increases. The watermark generation,
watermark embedding, and watermark extraction for
detection and authentication are three stages of digital
watermarking system .A watermarking scheme can be
classified into non-blind or blind watermarking basedonthe
method used to detect watermark. A non-blind
watermarking requires the original image to extract the
watermark, while a blind watermarking does not.
Watermark security refers to its resistance against
unauthorized detecting and decoding, while robustness
refers to a watermark’s resistance against processing, such
as filtering, geometrical transforms and compression. A
study on watermarking research shows that many
watermarking schemes give consideration to robustness
more than security. However, a robust watermark is not
enough to accomplish protection because the range of
hostile attacks is not limited to common processing and
distortions. Therefore, robustness and security should be
considered in a watermarking system. Spread-spectrum
technique is a popular approach to achieve the robust and
secure watermarking for multimedia content. A spread-
spectrum system encodes data in a chosen binary sequence
that appears like noise to an outsider but can be recognized
by a receiver with an appropriate key. A non-blind
watermarking might be less applicable, because when
watermark detection is required, the original imagemaynot
be available. To achieve security, the watermarking scheme
is constructed based on chaotic maps. A chaotic map is
function that exhibits some sorts of chaotic behaviour. The
feature of chaos in the information hiding is its sensitivity to
initial conditions. These characteristics make chaotic maps
excellent candidates for securing watermarks.
4. COMPARISON OF IMAGE AUTHENTICATION
METHODS.
Table -1: Comparision of image authentication methods
Techniques Advantages Disadvantages
Cryptographic
authentication
high tamper
detection
hash functions
are very sensitive
Robust image
hashing
authentication
Hashes produced
are robust;
collision
probability
between hashes
of different
images is very
low.
Large area
cropping
Watermarking
authentication
Image tampering
detection,
embedding
watermarks is
easy and identify
the author of
copyright work
Cannot locate
where the images
have been
tampered.
5. CONCLUSION
In multimedia era the digital information is sent across the
internet can potentially be intercepted by third party other
than the intended recipient. Therefore digital information
requires confidentiality, security service,authenticationand
determination of malicious activities. Literature Survey
concludes that watermarking scheme constructed based on
chaotic maps is more efficient compared to other methods,
because Chaos in information hiding are sensitive to initial
conditions and the outspreading of orbits over the entire
space. These characteristics make chaotic maps excellent
candidates for watermarking and encryption.
REFERENCES
[1] Y. Lei, Y.Wang, and J. Huang, “Robust image hash in
Radom transform domain for authentication,” IEEE Signal
Process. Image Communication.vol.26, no.6, pp.280288,
2011.
[2] V.Monga, M.K.Mihcak, “Robust and secure Imagehashing
via nonnegative matrix factorizations,” IEEE Trans. Inf.
Forensics Security, vol. 2, no. 3, pp.376– 390, Sep.2007.
[3] M. M. Yeung and F. Mintzer, "An invisible watermarking
technique for image verification," inProc.IEEEInternational](https://image.slidesharecdn.com/irjet-v5i12321-190108063825/85/IRJET-A-Study-on-Image-Authentication-Methods-2-320.jpg)
![International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 05 Issue: 12 | Dec 2018 www.irjet.net p-ISSN: 2395-0072
© 2018, IRJET | Impact Factor value: 7.211 | ISO 9001:2008 Certified Journal | Page 1721
Conference on Image Processing, VOL.2, pp. 680-683,
October 1997.
[4] C.Y. Lin and S.F. Chang, "Generating robust digital
signature for image/video authentication," in Proc.
Multimedia and Security Workshop at ACM Multimedia '98,
Bristol, UK, September 1998.
[5] C Yu, X Zhang “Watermark embedding in binary images
for authentication”, IEEE Trans. Signal Processing, VOL.01,
no.07, pp.865-868, September. 2004.
[6] H. Yang and A. C. Kot, “Binary image authentication with
tampering localization by embedding cryptographic
signature and block identifier,” IEEE Signal Processing
Letters, VOL. 13, Dec. 2006
[7] Yan Zhao, Shuozhong Wang, Guorui Feng, Zhenjun Tang,
”A Robust Image Hashing Method Based on Zernike
Moments”,IEEE Journal of Computational Information
Systems6:3 717-725,2011.
[8] Yan Zhao, Shuozhong Wang, Xinpeng Zhang, and Heng
Yao, Member, IEEE“Robust Hashing For Image
Authentication Using Zernike Moments And Local Features”
IEEE Transactions On Information Forensics And Security,
Vol.8, No. 1, January 2013.
[9] Dachselt F and Schwarz. W. “Chaos and Cryptography.
Circuits and Systems. Fundamental Theory and
Applications”, IEEE Transactions, 48(12), pp.1498-1509,
DOI- 10.1109/TCSI.2001.972857, 2001.](https://image.slidesharecdn.com/irjet-v5i12321-190108063825/85/IRJET-A-Study-on-Image-Authentication-Methods-3-320.jpg)
IRJET- A Study on Image Authentication Methods
- 1. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 05 Issue: 12 | Dec 2018 www.irjet.net p-ISSN: 2395-0072 © 2018, IRJET | Impact Factor value: 7.211 | ISO 9001:2008 Certified Journal | Page 1719 A Study on Image Authentication Methods Marakumbi Prakash R1, Jayashree V. Khanapuri2 1Asst.Professor, E&C Dept., Tontadarya College of Engineering, Gadag, Karnataka, INDIA 2Professor, E&T Dept.,K J Somaiya Institute of Engineering and Information Technology, Mumbai, Maharashtra, INDIA ----------------------------------------------------------------------***--------------------------------------------------------------------- Abstract - Authentication plays an important role in protecting image against unauthorized access. Digitalimages are transmitted over insecure channels such as the internet. Images must be protected against attempts to manipulate them; such manipulation could tamper the decisions based on these images. To protect the authenticity of images several methods have been proposed. Image authentication methods have gained attention due to their significance in the areas of multimedia communications and multimedia networking applications. The existing image authentication methods are cryptographic authentication, robust image hashing authentication and watermarking authentication. The aim of this paper is to present different methods in authentication of images in multimedia applications. Keywords: Image Authentication, Watermarking, Cryptography, Hashing 1. INTRODUCTION Authentication methods provide a means of ensuring the integrity of an image. Therefore there is need to protect these images against various attempts to manipulate them and it is important to make an effective method to solve image authentication problem that is ensuring the integrity of an image. Due to increase in the multimedia applications, image authentication techniques have gained attention. The existing image authentication methods are watermarking, cryptography and robust image hashing method. Digital watermarking is the science and art of embedding copyright information called watermarks in the files. Cryptography includes encryptionanddecryptiontotransferdocumentsor images. Robust image hashing is based onrotation-invariant moments that can effectively catch importantinformation in an image. 2. REQUIREMENTS OF IMAGE AUTHENTICATION The following are the requirements of image authentication: Sensitivity: The authentication system should beable to detect any content manipulation. Robustness: The authentication system musttolerate content preserving manipulations. Localization: The authenticationsystemshouldlocate the image regions that have been modified. Recovery: The authentication system should completely restore the image regions that were altered. Security: The authentication system must protect the authentication data against anyfalsificationattempts. 3. IMAGE AUTHENTICATION METHODS Various image authentication methods are being used to protect the images and information, and their communications. Virtual private network, firewall, encryption, cryptographic hash function such as digital signature, machine authentication code, manipulation detection code, and perceptual hashing are the examples of image authentication methods. Limitations of existing security like Firewall and virtual private network only protect the information up to the point of the internal networks. Encryption is an efficient tool for secure transmission, but when the sensitive data is decrypted, the information is not protected anymore. The drawback of cryptographic hash function is that it cannot locate where the images have been tampered. The major methods for authentication of an image are cryptographic, robust image hashing and watermarking authentication. 3.1 Cryptographic authentication Cryptography provides for secure communication in the presence of malicious third-parties known as adversaries. Encryption and Decryption are the two functions of cryptography. Traditional cryptography algorithms exhibit satisfactory results for image authentication with high tamper detection. Localization performances are not good but for some applications the method may be acceptable. Due to the sensitivity of the hash function, a small modification in binary image data causeschanges.Theimage is said to be manipulated even when one bit of this image is changed; this is very severe for most of the applications.The advantage of Cryptographic based authentication is conventional cryptography show satisfying results for authentication of an image with high tamper detection. The disadvantages of Cryptographic authentication are hash functions are very sensitive, localization performances are not very good.
- 2. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 05 Issue: 12 | Dec 2018 www.irjet.net p-ISSN: 2395-0072 © 2018, IRJET | Impact Factor value: 7.211 | ISO 9001:2008 Certified Journal | Page 1720 3.2 Robust image hashing authentication Image hashing is used to identify the duplicate copies of the original images. A robust hashing technique is used for detecting imageforgery,colourmodification,andforlocating the forged area. The robust hashing method is based on Random Transform performed on input image to obtain the projections in various orientations. The insignificant coefficients are removed. The invariant moments are calculated. DFT is performed on the invariant moments. Finally, the image hash value H is defined as the normalized and quantized version of the significant DFT coefficientsand then the hash bits are generated. Most of the images hashing algorithms have their disadvantages in getting the desirable performance against a particular image processing attack. The disadvantage of robust image hashing is large area cropping. 3.3 Watermarking authentication Digital watermarking involves embedding information into digital multimedia content such that the information can later can be extracted or detected by the recipient for different purposes such as control and copy prevention.The process of watermarking has to be resilient against tampering attacks, keeping the content of a watermark readable in order to be recognizable when extracted by the recipient. The significant features of watermarking system are robustness and fidelity. The size of the embedded information has to be considered. The data becomes less robust as its size increases. The watermark generation, watermark embedding, and watermark extraction for detection and authentication are three stages of digital watermarking system .A watermarking scheme can be classified into non-blind or blind watermarking basedonthe method used to detect watermark. A non-blind watermarking requires the original image to extract the watermark, while a blind watermarking does not. Watermark security refers to its resistance against unauthorized detecting and decoding, while robustness refers to a watermark’s resistance against processing, such as filtering, geometrical transforms and compression. A study on watermarking research shows that many watermarking schemes give consideration to robustness more than security. However, a robust watermark is not enough to accomplish protection because the range of hostile attacks is not limited to common processing and distortions. Therefore, robustness and security should be considered in a watermarking system. Spread-spectrum technique is a popular approach to achieve the robust and secure watermarking for multimedia content. A spread- spectrum system encodes data in a chosen binary sequence that appears like noise to an outsider but can be recognized by a receiver with an appropriate key. A non-blind watermarking might be less applicable, because when watermark detection is required, the original imagemaynot be available. To achieve security, the watermarking scheme is constructed based on chaotic maps. A chaotic map is function that exhibits some sorts of chaotic behaviour. The feature of chaos in the information hiding is its sensitivity to initial conditions. These characteristics make chaotic maps excellent candidates for securing watermarks. 4. COMPARISON OF IMAGE AUTHENTICATION METHODS. Table -1: Comparision of image authentication methods Techniques Advantages Disadvantages Cryptographic authentication high tamper detection hash functions are very sensitive Robust image hashing authentication Hashes produced are robust; collision probability between hashes of different images is very low. Large area cropping Watermarking authentication Image tampering detection, embedding watermarks is easy and identify the author of copyright work Cannot locate where the images have been tampered. 5. CONCLUSION In multimedia era the digital information is sent across the internet can potentially be intercepted by third party other than the intended recipient. Therefore digital information requires confidentiality, security service,authenticationand determination of malicious activities. Literature Survey concludes that watermarking scheme constructed based on chaotic maps is more efficient compared to other methods, because Chaos in information hiding are sensitive to initial conditions and the outspreading of orbits over the entire space. These characteristics make chaotic maps excellent candidates for watermarking and encryption. REFERENCES [1] Y. Lei, Y.Wang, and J. Huang, “Robust image hash in Radom transform domain for authentication,” IEEE Signal Process. Image Communication.vol.26, no.6, pp.280288, 2011. [2] V.Monga, M.K.Mihcak, “Robust and secure Imagehashing via nonnegative matrix factorizations,” IEEE Trans. Inf. Forensics Security, vol. 2, no. 3, pp.376– 390, Sep.2007. [3] M. M. Yeung and F. Mintzer, "An invisible watermarking technique for image verification," inProc.IEEEInternational
- 3. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 05 Issue: 12 | Dec 2018 www.irjet.net p-ISSN: 2395-0072 © 2018, IRJET | Impact Factor value: 7.211 | ISO 9001:2008 Certified Journal | Page 1721 Conference on Image Processing, VOL.2, pp. 680-683, October 1997. [4] C.Y. Lin and S.F. Chang, "Generating robust digital signature for image/video authentication," in Proc. Multimedia and Security Workshop at ACM Multimedia '98, Bristol, UK, September 1998. [5] C Yu, X Zhang “Watermark embedding in binary images for authentication”, IEEE Trans. Signal Processing, VOL.01, no.07, pp.865-868, September. 2004. [6] H. Yang and A. C. Kot, “Binary image authentication with tampering localization by embedding cryptographic signature and block identifier,” IEEE Signal Processing Letters, VOL. 13, Dec. 2006 [7] Yan Zhao, Shuozhong Wang, Guorui Feng, Zhenjun Tang, ”A Robust Image Hashing Method Based on Zernike Moments”,IEEE Journal of Computational Information Systems6:3 717-725,2011. [8] Yan Zhao, Shuozhong Wang, Xinpeng Zhang, and Heng Yao, Member, IEEE“Robust Hashing For Image Authentication Using Zernike Moments And Local Features” IEEE Transactions On Information Forensics And Security, Vol.8, No. 1, January 2013. [9] Dachselt F and Schwarz. W. “Chaos and Cryptography. Circuits and Systems. Fundamental Theory and Applications”, IEEE Transactions, 48(12), pp.1498-1509, DOI- 10.1109/TCSI.2001.972857, 2001.