K. B. Raja

High Capacity Lossless Secure Image Steganography using Wavelets

The modern steganography presents a challenging task of embedding data that should be imperceptible to the human visual system (HVS) and also escape the detection of powerful machine vision of computers. In this paper we present a high capacity, lossless, secure wavelet steganographic algorithm in which payload bitstream is encrypted and embedded into the wavelet coefficients of the cover image to derive a stego-image. The payload is embedded in the approximation band of the wavelet domain that increases its robustness. It is observed through simulations that mean square error (MSE), mean absolute error (MAE), bit error rate (BER) and histogram analysis that the pay load is retrieved without any errors and its performance is better than the earlier insignificant coefficient replacement (ICR) technique.

Genetic algorithm based steganography using wavelets

Steganography has long been a means of secure communication. Security is achieved by camouflaging the secret message. In this paper, we present a Genetic Algorithm based Steganography using Discrete Cosine Transforms(GASDCT) and Genetic Algorithm based Steganography using DiscreteWavelets Transform(GASDWT). In our approach, the Discrete Cosine Transform and Discrete Wavelet Transform are applied to the payload. Genetic Algorithm is used to generate many stego-images based on Fitness functions; one of these which give least statistical evidence of payload is selected as the best stego image to be communicated to the destination. It is observed that GASDWT has an improvement in Bit Error Rate(BER), Peak Signal to Noise Ratio(PSNR) and embedding capacity as compared to GASDCT.

Offline signature authentication using cross-validated graph matching

The biometric system is used to identify a person depending on his physiological or behavioral characteristics. Signature verification is a commonly accepted biometric method and is widely used for banking transactions. In this paper, we propose Offline Signature Authentication using Cross-validated Graph Matching (OSACGM) algorithm. The signatures are pre-processed in which signature extraction method is used to obtain high resolution for smaller normalization box. The similarity measure between two signatures in the database is determined by (i) constructing a bipartite graph G, (ii) obtaining complete matching in G and (iii) finding minimum Euclidean distance by Hungarian method. An optimum decision threshold value is determined using Cross-validation technique to select reference signatures. The test feature is extracted from the given test signature by pre-processing. Then the test feature is compared with the threshold value to authenticate the test signature. Compared to the existing algorithm, our algorithm gives better Equal Error Rate (EER) for skilled and random forgeries.

Iris feature extraction using directional filter bank for personal identification

Iris Recognition is a biometric tool, which has great emphasis in both research and practical applications. In this paper an Iris Recognition System using Directional Filter Bank (IRSDFB) is proposed. The normalized Iris is fragmented into three regions. The most distinctive features of the region nearer to the pupil are encoded to form the feature vector using Directional Filter Bank, instead of considering the entire Iris including the occluded portion. Therefore the Iris images captured with lesser cooperation can also be verified successfully. The decidability index of IRSDFB model is more compared to the existing algorithm using Gabor filter bank.

DBC based Face Recognition using DWT

The applications using face biometric has proved its reliability in last decade. In this paper, we propose DBC based Face Recognition using DWT (DBC- FR) model. The Poly-U Near Infra Red (NIR) database images are scanned and cropped to get only the face part in pre-processing. The face part is resized to 100*100 and DWT is applied to derive LL, LH, HL and HH subbands. The LL subband of size 50*50 is converted into 100 cells with 5*5 dimention of each cell. The Directional Binary Code (DBC) is applied on each 5*5 cell to derive 100 features. The Euclidian distance measure is used to compare the features of test image and database images. The proposed algorithm render better percentage recognition rate compared to the existing algorithm.

Fault Tolerant QoS Adaptive Clustering for Wireless Sensor Networks

This paper proposes and analyzes an Energy Efficient Fault Tolerant QoS Adaptive Clustering Algorithm (FTQAC) for Wireless sensor networks suitable to support real-time traffic. The protocol achieves fault tolerance and energy efficiency through a dual cluster head mechanism and guarantees the desired QoS by including delay and bandwidth parameters in the route selection process. Simulation results indicate that FTQAC reduces overall energy consumption and improves network lifetime while maintaining required QoS.

Extracting hidden image using histogram, DFT and SVM

In covert communication, Information hiding is rapidly gaining momentum. There are many sophisticated techniques being developed in steganography. There is a need of universal method to detect hidden image. We have proposed a Universal method to detect hidden message using Histogram, Discrete Fourier Transform and SVM (UDHDS). When compared to cover image stego image has irregular statistical characteristics. one class SVM is trained by these statistical features which are generated Using Histogram and DFT to discriminate the cover and stego image. The number of statistical features is less in UDHDS Algorithm when compared to the existing algorithm and found to be more efficient.

Mantissa replacement steganography using LWT

Steganography is an authenticated technique for maintaining secrecy of embedded data. The novel concept of replacing mantissa part of cover image by the generated mantissa part of payload is proposed for higher capacity and security. The Lifting wavelet Transform (LWT) is applied on both cover image and payload of sizes a * a and 3a * 2a respectively. The mantissa values of Vertical band (CV), Horizontal band (CH) and diagonal band (CD) of cover image are removed to convert into real values. The approximation band of payload is considered and the odd column element values and even column element values are divided by 300 and 30000 respectively to generate only mantissa part of payload. The modified odd and even column vector pairs are added element by element to form one resultant vector. The column vector elements of cover image and resultant column vector elements of payload are added to generate stego object column vector elements corresponding to vertical, horizontal and diagonal elements. The inverse LWT is applied to generate stego image.

Recognizing Human Faces with Tilt

Issues related to realtime face recognition are perpetual even with many existing approaches. Generalizing these issues is tedious over different applications. In this paper, the real time issues such as tilt or rotation variation and few samples problem for face recognition are addressed and proposed an efficient method. In preprocessing, an edge detection method using Robert`s operator is utilized to identify facial borders for cropping purpose. The query images are axially tilted for different degrees of rotation. Both database and test images are segmented into one hundred fragments of 5 * 5 size each. Four different matrix characteristics are derived for each divided part of the image. Corresponding attributes are added to yield features related to final matrix. Final one hundred facial attributes are obtained by fusing diagonal features with one hundred features of matrix. Euclidean distance between the final attributes of gallery and query images is computed. The results on Yale dataset has superior performance compared to the existing different approaches and it is convincing over the dataset created.

PCVOS: Principal component variances based off-line signature verification

Offline signature verification system is widely used as a behavioral biometric for identifying a person. This behavioral biometric trait is a challenge in designing the system that has to counter intrapersonal and interpersonal variations. In this paper, we propose a novel technique PCVOS: Principal Component Variances based Off-line Signature Verification on two critical parameters viz., the Pixel Density (PD) and the Centre of Gravity (CoG) distance. It consists of two parallel processes, namely Signature training which involves extraction of features from the samples of database and Test signature analysis which performs extraction of features from the test samples. The trained values from the database are compared with the features of the test signature using Principal Component Analysis (PCA). The PCVOS algorithm shows a notable improvement over the algorithms in [21], [22] and [23].