#### DMCA

## ABSTRACT Data Encryption and Hashing Schemes for Multimedia Protection (2012)

### Citations

3332 | Handbook of applied cryptography
- MENEZES, OORSCHOT, et al.
(Show Context)
Citation Context ...fferent applications. For example, DVD and Pay-TV, which exist for the purpose of profit and revenue, require a higher level than video surveillance cameras, which basically exist to provide security =-=[2]-=-. 1.3.2 SPEED Images and multimedia, compared to text, constitute a major part of internet traffic. Thus, image and multimedia encryption algorithms should be designed with more emphasis on execution ... |

678 |
On the shortest spanning subtree of a graph and the traveling salesman problem,
- Kruskal
- 1956
(Show Context)
Citation Context ...s can be used to calculate MST, such as: • Boruvka’s algorithm • Prim’s algorithm • Reverse-Delete algorithm • Kruskal’s algorithm Kruscal’s algorithm is a fast method and is described in Algorithm 2 =-=[35]-=-. 3.2.6 ENTROPY OF A GRAPH Entropy is the measure of unpredictability in a random process. For example, a fair coin has higher entropy than an unfair coin, because a series of tosses with a fair coin ... |

401 | METIS: A software package for partitioning unstructured graphs, partitioning meshes, and computing fill-reducing orderings of sparse matrices”,
- Karypis, Kumar
- 1997
(Show Context)
Citation Context ...r process and is proportional toO(m3). Thus, we need partition a mesh to smaller sub-meshes to increase the hash generation speed. The implementation of partitioning algorithm is based onMeTiS method =-=[40, 41, 42]-=-. The proposed method is summarized in Algorithm 3 and Fig. 3.7. Figs. 3.20-3.21 depict the 3D shark and dragon models and their eight corresponding sub-meshes. The entropy of each partition is writte... |

119 |
Applications of entropic spanning graphs.
- Hero, Ma, et al.
- 2002
(Show Context)
Citation Context ...̂α(V) = 1 1− α [ L∗(V) βmα − 1 ] (3.21) where L∗(V) is the total length of the minimum spanning tree [34]. α is referred to as an entropic index, and β is a constant playing a role of bias correction =-=[38]-=-. To calculate the minimal spanning tree we use Kruskal’s algorithm. 3.3 PROPOSED METHOD The proposed method uses the Laplace-Beltrami matrix eigen-decomposition and Tsallis entropy of the spanning tr... |

105 |
Image and Video Compression for Multimedia Engineering,
- Shi, Sun
- 2000
(Show Context)
Citation Context ...image or video FIGURE 1.2: Image/video compress/uncompress block diagram. or video exactly matches the original. However, in lossless methods, the compression ratio is not as good as in lossy methods =-=[3]-=-. Some of the lossless image compression methods are: • Prediction based methods [4] • DPCM and Predictive Coding • Entropy encoding Some lossy compression methods for image compression are: • Fractal... |

95 |
Sparse multidimensional representation using shearlets. Wavelets XI.,
- Labate, Lim, et al.
- 2005
(Show Context)
Citation Context ...es of surfaces [16]. Shearlet, on the other hand, is more efficient in modeling the geometrical property of an image due to its anisotropic nature. There are different ways to define Shearlet and DST =-=[6, 16, 17]-=-. 2.4.2 CONTINUOUS SHEARLET TRANSFORM The Fourier transform reveals only the frequency attributes of an image. The wavelet transform provides a powerful insight into an image’s spatial and frequency c... |

81 |
Automatic mesh partitioning.
- Miller, Teng, et al.
- 1993
(Show Context)
Citation Context ...r process and is proportional toO(m3). Thus, we need partition a mesh to smaller sub-meshes to increase the hash generation speed. The implementation of partitioning algorithm is based onMeTiS method =-=[40, 41, 42]-=-. The proposed method is summarized in Algorithm 3 and Fig. 3.7. Figs. 3.20-3.21 depict the 3D shark and dragon models and their eight corresponding sub-meshes. The entropy of each partition is writte... |

72 |
Image and Video Encryption: From Digital Rights Management to Secured Personal Communication, ser.
- Uhl, Pommer
- 2005
(Show Context)
Citation Context ... and non-visual data like patient records is a private and confidential property of its owners. There is, therefore, an increasing demand for data security, i.e. privacy, integrity and authentication =-=[1]-=-. Techniques like watermarking and image encryption are used to provide authentication, confidentiality, and privacy to images and other types of visual data like video and 3D objects. In this chapter... |

62 | A wavelet-based analysis of fractal image compression,
- Davis
- 1998
(Show Context)
Citation Context ... lossless image compression methods are: • Prediction based methods [4] • DPCM and Predictive Coding • Entropy encoding Some lossy compression methods for image compression are: • Fractal compression =-=[5]-=- • Transform coding • Chroma subsampling The transform coding is the most commonly used lossy compression method. It involves the application of a Fourier-related transform such as the Discrete Fourie... |

39 |
Image encryption for secure internet multimedia applications,
- Dang, Chau
- 2000
(Show Context)
Citation Context ...encrypted data. In data coding, encrypted data is sent through an insecure channel. Data Encryption Standard (DES) and Advanced Encryption Standard (AES) 16 are widely-used methods of data encryption =-=[8]-=-. Channel encryption and data encryption can be combined to improve security, but this process requires more resources and more expensive hardware. Although the bandwidth price has decreased considera... |

34 |
Evans,“Perceptual image hashing via feature points: Performance evaluation and tradeoffs”,
- Monga, L
- 2006
(Show Context)
Citation Context ...ld have two different hash values. Conventional hash functions like MD5 or SHA-1 are extremely sensitive to input variation. i.e. 1 bit change in the input message changes the hash value dramatically =-=[28]-=-. Perceptual image hash or perceptual 3D mesh hash functions are different. They create similar hash values for two 3D objects which look the same 38 to the human eye. These perceptual hash algorithms... |

29 | Spectral methods for mesh processing and analysis,” in EG State-of-the-art Report,
- Zhang, Kaick, et al.
- 2007
(Show Context)
Citation Context ... learning has become more widespread. Generally speaking, a spectral method calculates the eigenvalues and eigenvectors corresponding to a linear operator and try to use them to solve a given problem =-=[33]-=-. This method can be used in variety of problems, including mesh parametrization, mesh authorization, and mesh compression. The primary motivation of this chapter is to propose a method for 3D mesh au... |

27 | Geometric separators for finite-element meshes.
- Miller, Teng, et al.
- 1998
(Show Context)
Citation Context ...r process and is proportional toO(m3). Thus, we need partition a mesh to smaller sub-meshes to increase the hash generation speed. The implementation of partitioning algorithm is based onMeTiS method =-=[40, 41, 42]-=-. The proposed method is summarized in Algorithm 3 and Fig. 3.7. Figs. 3.20-3.21 depict the 3D shark and dragon models and their eight corresponding sub-meshes. The entropy of each partition is writte... |

25 |
The discrete shearlet transform: a new directional transform and compactly supported shearlet frames.
- Lim
- 2010
(Show Context)
Citation Context ...tify the location of an image edge, but also the geometrical property of the image. Wavelet is isotropic, and is therefore able to capture the geometric regularity along the singularities of surfaces =-=[16]-=-. Shearlet, on the other hand, is more efficient in modeling the geometrical property of an image due to its anisotropic nature. There are different ways to define Shearlet and DST [6, 16, 17]. 2.4.2 ... |

18 |
Shearlet-based total variation diffusion for denoising
- Easley, Labate, et al.
(Show Context)
Citation Context ... ψa,s,t equation, we can obtain ψj,l,k =| detA0 | j 2 ψ(Bl0A j 0x− k) (1.34) for j, l ∈ Z, k ∈ Z2. The discrete shearlet is able to deal with multi-dimensional functions as its continuous counterpart =-=[7]-=-. 1.8.3 APPLICATIONS OF SHEARLET TRANSFORM There are various fields where the shearlet is used. Generally, all applications of the DWT can employ the DST too. The following list shows some popular usa... |

15 | Methods of attacking chaotic encryption and countermeasures”,
- Sobhy, Shehata
- 2001
(Show Context)
Citation Context ...rturbed chaotic PWLCM maps are used in the multiple substitution/permutation rounds. Although the chaotic maps can create high quality random numbers, they are more prone to different type of attacks =-=[21]-=-. 2.3 ADVANCED ENCRYPTION STANDARD (AES) AES is a cryptography symmetric key encryption standard that was adopted for commercial encryption usages [19, 20]. It is a new version of symmetric key encryp... |

12 | Normalized cuts and image segmentation - Jianbo, Malik - 2000 |

11 |
Extended Instructions for The AES Cryptography and Their Efficient Implementation
- Nadehara, Ikekawa, et al.
- 2004
(Show Context)
Citation Context ...they are more prone to different type of attacks [21]. 2.3 ADVANCED ENCRYPTION STANDARD (AES) AES is a cryptography symmetric key encryption standard that was adopted for commercial encryption usages =-=[19, 20]-=-. It is a new version of symmetric key encryption/decryption protocol that supersedes DES. The major weakness of DES or Triple DES (3DES) is low encryption speed, in software as well as hardware imple... |

10 |
Possible generalizations of Boltzman-Gibbs statistics.
- Tsallis
- 1988
(Show Context)
Citation Context ... entropy is the simple addition (sum) of the entropy of each subsystem and the correlations between subsystems are not accounted for. To overcome this weakness, we use the Tsallis entropy of a system =-=[37]-=-. Tsallis entropy of a discrete distribution is defined as Ĥα(X) = 1 α− 1 ( 1− n∑ i=1 pαi ) (3.17) where α is a real parameter. When α is close to 1, the normal Boltzmann-Gibbs entropy or Shannon-Ré... |

7 | Near lossless data compression onboard a hyperspectral satellite - Bergeron, Cunningham, et al. - 2006 |

7 | Robust image hashing via nonnegative matrix factorizations,” in
- Monga, Mihçak
- 2006
(Show Context)
Citation Context ... This projection process is non-invertible and non-linear, so that different hashes can be extracted from the same image if random permutations of the input are used. Another method was introduced in =-=[25]-=-, which employs the use of non-negative matrix factorization (NMF) for robust image hashing. The basic idea is to employ the image matrix to create an algorithm for randomized dimensionality reduction... |

6 |
On Rényi and Tsallis entropies and divergences for exponential families
- Nielsen, Nock
- 1105
(Show Context)
Citation Context ...nd while where X is a discrete random variable with possible values {x1, x2, ..., xn}, p1, p2, ..., pn are the probabilities of x1, x2, ..., xn respectively, and α ≥ 0, α = 1 is the order of entropy =-=[36]-=-. If α → 1, then Shannon entropy is achieved Ĥα(X) = − n∑ i=1 pi log2 pi (3.16) As a generalization of Boltzmann-Gibbs entropy, Tsallis entropy has recently gained a great deal of interest in statist... |

5 | From Wavelets to Shearlets and back again
- Kutyniok, Sauer
- 2008
(Show Context)
Citation Context ...)) (2.8) of the shearlet group S = R+ × R × R2. The location parameter t can precisely detect the location of singularities, and the shear parameter s shows the perpendicular direction to singularity =-=[18]-=-. (a) (b) FIGURE 2.2: Geometric transformation in shearlet: (a) anisotropic dilation (scaling) parameter a, (b) shear parameter s. 2.4.3 DISCRETE SHEARLET TRANSFORM (DST) By sampling continuous Shearl... |

4 |
A.K Tiwari, “A survey on lossless image compression methods
- Shukla, Alwani
- 2010
(Show Context)
Citation Context ...b : R → R be b(ω) = ⎧⎪⎪⎪⎪⎪⎨⎪⎪⎪⎪⎪⎩ sin((π/2)v(|ω| − 1)), 1 ≤ |ω| ≤ 2, cos((π/2)v(|ω|/2 − 1)), 2 < |ω| ≤ 4, 0, else. (1.20) The function b is symmetric, positive, real function, and supp(b) = [−4,−1] ∪ =-=[1, 4]-=-. Fig. 1.5 shows the v(x) and b(ω) functions. We define ψ1 via its Fourier transform as ψ̂1(ω) = √ b2(ω) + b2(2ω) (1.21) The second function ψ2 is defined again via its Fourier transform as ψ̂2(ω) = ⎧... |

4 | Fast finite shearlet transform: a tutorial. arXiv preprint arXiv:1202.1773
- Häuser, Steidl
- 2014
(Show Context)
Citation Context ...es of surfaces [16]. Shearlet, on the other hand, is more efficient in modeling the geometrical property of an image due to its anisotropic nature. There are different ways to define Shearlet and DST =-=[6, 16, 17]-=-. 2.4.2 CONTINUOUS SHEARLET TRANSFORM The Fourier transform reveals only the frequency attributes of an image. The wavelet transform provides a powerful insight into an image’s spatial and frequency c... |

4 |
A new chaos-based cryptosystem for secure transmitted images
- Awad
(Show Context)
Citation Context ... shearlet transform in image approximation and compression. 1.9 ENTROPY Entropy is the statistical measure of uncertainty in a random variable. The Shannon entropy of a gray scale image is defined as =-=[14]-=- H = − K∑ i=1 pi log2 pi (1.35) where pi is the probability of color with value i, andK is the total number of different colors (K = 256 for a gray scale image). When a gray scale message is ideally e... |

4 |
Spanning Trees and Optimization problems, Chapman & Hall/CRC
- wu, Chao
- 2004
(Show Context)
Citation Context ...j], i = j, along with a list of edge adjacency relations. Fig. 3.5 shows a 4-vertex complete graph and Fig. 3.6 depicts some of its spanning trees. FIGURE 3.5: A 4-vertex complete graph. As shown in =-=[34]-=- a graph with n vertices has Kn distinct spanning trees. Kn = n n−2 (3.14) For example, a graph with n = 6 vertices has 1296 distinct spanning trees. A weighted graph associates a label (a real number... |

3 | Randomness enhancement using digitalized modified logistic map, - Chen, Hwang, et al. - 2010 |

2 |
A new compound chaos encryption algorithm for digital images
- Jun, Jinping, et al.
- 2010
(Show Context)
Citation Context ...ssed in [11, 10]. More secure methods should, therefore, be chosen. 17 2.2.2 CHAOS BASED ENCRYPTION In recent years, a slew of methods were proposed to improve encryption using chaos-based algorithms =-=[12, 13, 14]-=-. Awad in [14] proposed four chaos based cryptosystems. Using these algorithms, a two-dimensional (2D) chaotic map is used to shuffle the pixels’ locations in an image. Then, two perturbed chaotic PWL... |

2 |
Orthogonal Bases and the QR Algorithm
- Olver
- 2008
(Show Context)
Citation Context ...s theorem, roots of a polynomial equation of order 5th and higher cannot be expressed simply using nth roots. The iterative algorithms like • Power method [30] • Arnoldi iteration [31] • QR algorithm =-=[32]-=- are often employed to compute eigenvalues. 3.2.3 SPECTRAL ANALYSIS OF A 3D MESH Most spectral methods have a basic framework that can be summarized in three steps [33] • A matrix L = (Lij) is defined... |

1 | On the hardware design for DES cipher in tamper resistant devices against differential fault analysis - Wang, Laih, et al. - 2000 |

1 |
An attack against DES based on the relationship
- Cao, Liu
- 2010
(Show Context)
Citation Context ...e for image compression. However, DES is not a secure method by today’s security standards, and there are numerous ways to attack it, such as the improved Davies’ attack, and the attacks discussed in =-=[11, 10]-=-. More secure methods should, therefore, be chosen. 17 2.2.2 CHAOS BASED ENCRYPTION In recent years, a slew of methods were proposed to improve encryption using chaos-based algorithms [12, 13, 14]. Aw... |

1 |
Successful attack on an FPGA-based WDDL DES cryptoprocessor without place and route constraints
- Sauvage, Guilley, et al.
- 2009
(Show Context)
Citation Context ...e for image compression. However, DES is not a secure method by today’s security standards, and there are numerous ways to attack it, such as the improved Davies’ attack, and the attacks discussed in =-=[11, 10]-=-. More secure methods should, therefore, be chosen. 17 2.2.2 CHAOS BASED ENCRYPTION In recent years, a slew of methods were proposed to improve encryption using chaos-based algorithms [12, 13, 14]. Aw... |

1 |
An encryption method based on dual-chaos system
- Wang, Ma, et al.
- 2009
(Show Context)
Citation Context ...ssed in [11, 10]. More secure methods should, therefore, be chosen. 17 2.2.2 CHAOS BASED ENCRYPTION In recent years, a slew of methods were proposed to improve encryption using chaos-based algorithms =-=[12, 13, 14]-=-. Awad in [14] proposed four chaos based cryptosystems. Using these algorithms, a two-dimensional (2D) chaotic map is used to shuffle the pixels’ locations in an image. Then, two perturbed chaotic PWL... |

1 |
Robust 3D watermarking based on geometry image
- Wang, Zheng, et al.
- 2008
(Show Context)
Citation Context ...s to the complexity of the issue. To tackle this problem, 3D watermarking and hash functions can be helpful. In 3D watermarking, some data are hidden in a 3D object, and can be used for authorization =-=[22]-=-. The cryptographic hash functions can also be used to control the integrity and the authentication of data. A cryptography hash function is a map that takes an input with arbitrary length and generat... |

1 |
A new hash algorithm based on MQ problem and polymorphic cipher
- Wang, Zhang, et al.
- 2011
(Show Context)
Citation Context ... A cryptography hash function is a map that takes an input with arbitrary length and generates a fixed length output. Usually this output is a binary string, which is referred to as the hash of input =-=[23]-=-. By recalculating the hash value from underlying data and comparing it to the attached hash value, we can 35 verify the authenticity of data [39]. In [45], a robust method for image watermarking was ... |

1 | Robust image hashing using higher order spectral features - Chen, Chandran - 2010 |

1 |
DWT based robust image hashing algorithm
- Hu, Niu
- 2010
(Show Context)
Citation Context ...employ the image matrix to create an algorithm for randomized dimensionality reduction that retains the essence of the original image matrix, while resisting different attacks. Recently, Hu et al. in =-=[26]-=- used a new variation of the discrete wavelet transform (DWT) algorithm. First, this method extracts robust bits of High pass-Low pass (HL), Low pass-High pass (LH) and High pass-High pass (HH) sub-ba... |

1 |
A learning framework for robust hashing of face images
- Senel, Mihak
- 2010
(Show Context)
Citation Context ...cy scale. A a hash is then generated, based on these sub-bands. Other recent development includes the presentation of a two stage cascade of dimensionality reduction constructs for face image hashing =-=[27]-=-. The first stage aims to project the face image to a space where geometric distortions manifest approximately as additive noise using the NMF method. In the second stage, Oriented Principal Component... |

1 |
Natural power method for fast subspace tracking,” Eurographics STAR report
- Yingbo, Yong, et al.
- 1999
(Show Context)
Citation Context ...hods are employed because, according to Abel’s theorem, roots of a polynomial equation of order 5th and higher cannot be expressed simply using nth roots. The iterative algorithms like • Power method =-=[30]-=- • Arnoldi iteration [31] • QR algorithm [32] are often employed to compute eigenvalues. 3.2.3 SPECTRAL ANALYSIS OF A 3D MESH Most spectral methods have a basic framework that can be summarized in thr... |

1 |
A new method for accelerating Arnoldi algorithms for large scale eigenproblems
- Dookhitram, Boojhawon, et al.
- 2009
(Show Context)
Citation Context ..., according to Abel’s theorem, roots of a polynomial equation of order 5th and higher cannot be expressed simply using nth roots. The iterative algorithms like • Power method [30] • Arnoldi iteration =-=[31]-=- • QR algorithm [32] are often employed to compute eigenvalues. 3.2.3 SPECTRAL ANALYSIS OF A 3D MESH Most spectral methods have a basic framework that can be summarized in three steps [33] • A matrix ... |

1 |
Hamza , “Entropic hashing of 3D objects using LaplaceBeltrami operator
- Ghaderpanah, Abbas, et al.
- 2008
(Show Context)
Citation Context ...ression ratio, and increases the security of transmitted images over the Internet. ❏ In Chapter 3, we present a hashing technique for 3D models using spectral graph theory and entropic spanning trees =-=[39]-=-. The main idea is to partition a 3D triangle mesh into an ensemble of sub-meshes, then apply eigen-decomposition to the Laplace-Beltrami matrix of each sub-mesh, followed by the computation of the ha... |

1 | Image encryption in wavelet domain based on chaotic maps - Podoba, Giesl, et al. - 2009 |