Information Technology

In this paper, we propose an integrated face recognition system that is robust against facial expressions by combining information from the computed intra-person optical flow and the synthesized face image in a probabilistic framework. Making recognition more reliable under uncontrolled lighting conditions is one of the most important challenges for practical face recognition systems. We tackle this by combining the strengths of robust illumination normalization. Our experimental results show that the proposed system improves the accuracy of face recognition from expressional face images and lighting variations.

To secure data or information by a modified RSA cryptosystem based on ‘n’ prime. This is a new technique to provide maximum security for data over the network. It is involved encryption, decryption, and key generation. Prime number used in a modified RSA cryptosystem to provide security over the networks. In this technique we used ‘n’ prime number which is not easily breakable. ‘n’ prime numbers are not easily decompose. This technique provides more efficiency and reliability over the networks. In this paper we are used a modified RSA cryptosystem algorithm to handle ‘n’ prime numbers and provides security.

The problem of identifying approximately duplicate records in databases is an essential step for data cleaning and data integration processes. In the real world, entities have two or more representations in databases. Errors are introduced as the result of transcription errors, incomplete information, lack of standard formats, or any combination of these factors. Most of the existing approach has been depend on the generic or manual intervention is required to calculate the edit distance. None of the existing approach hasn’t focus on how many individual operations required for edit distance calculation, existing relied on total operation. In this paper, we present a thorough analysis of the literature on edit distance for duplicate detection. We cover similarity metrics that are commonly used to detect number of individual operation required to find a duplicate record, which includes the total required edit operation.

Bluetooth is a promising short range wireless personal area network technology and is on the verge of being ubiquitously deployed over a wide range of devices. The basic unit of a Bluetooth network is a centralized master-slave topology, namely a piconet, which can be easily extended into a multi-hop adhoc network called a scatternet. Scatternets increase Bluetooth’s usability multi-folds such that numerous applications may be built over them to unleash the potential of Bluetooth. When more than seven devices to be connected in a Bluetooth scatternet, bridge devices are used to connect two piconets into a scatternet. To deal with possible data transmissions between different piconets, the bridge device must switch to different masters frequently. This may cause traffic in network.Our work tries to explore a less popular but powerful applications of Bluetooth which can help extend the range of Bluetooth communication and provide a free infrastructure for communication and find the transaction delay in Bluetooth network. Our approach is to design a transaction system that can be used to emulate and test the path selection from plentiful paths available from source to the destination. At the same this it assist user to select the optimal path. Due to this innovative approach; in transaction system it cause less traffic and it resultant in to less delay. Our work also analyzes the end-to-end packet delays, cumulative delay for both intra- and inter-piconet traffic and energy consumption, in a Bluetooth scatternet. Finally, we conduct simulations by using eclipse and NS-2 to demonstrate the efficiency and effectiveness of the proposed transaction system.

The incredible growth of the information space, the hard competition between enterprises populating it, the need to collaborate and to share knowledge guiding to better decisions are factors that push Web Services providers to join each other into Web Services communities. Solutions proposed to tackle communities’ management lack of a formal and clear design which may alter the system evolving and maintenance. For these reasons, we propose to design and to implement a system able to manage Web Services communities using an agent oriented approach. As illustration, we design and implement a case tool for managing e-health care Web Services communities.

In the real world, most people are socially selfish; i.e., they are willing to forward packets with whom they have social ties but not to others, which varies with the strength of the social tie. A Social Selfishness Aware Routing (SSAR) algorithm to allow user selfishness and for having better routing performance in an efficient way is proposed. To select a forwarding node, SSAR considers both users’ willingness to forward and their contact based approaches, which results in a better forwarding method than purely contact-opportunities. SSAR also formulates the packet forwarding process as a Multiple Knapsack Problem with Assignment Restrictions (MKPAR) to satisfy user demands for selfishness and performance. Trace-based simulations show that SSAR allows users to maintain selfishness and it achieves better routing performance with low transmission cost.

In this paper, It examines the Face detection and recognition system using three fast CFV modules which are face skin verification module, face symmetry verification module, and eye template verification module. The three verification modules can eliminate the tilted faces, the backs of the head, and any other non-face moving objects in the video. Only the frontal face images are sent to face recognition engine. The frontal face detection reliability can be adjusted by simply setting the verification thresholds in the verification modules. Then three hybrid feature sets are applied to face recognition. Experiments demonstrated that the frontal face detection rate can be achieved as high as 95% in the low quality video images. The overall face recognition rate and reliability are increased at the same time using the proposed ensemble classifier in the system. An ensemble classifier scheme is proposed to congregate three individual Artificial Neural Network classifiers trained by the three hybrid feature sets.

The rapid development of network technologies has lead to vast multimedia data being transmitted over the networks. However, the data being transmitted can be tampered or attacked by some malicious attackers during transit. To provide security to multimedia data steganography is being used as a potential tool, where the secret data to be transmitted is embedded in a cover medium (Image) in order to avert the attackers. This paper presents a secure and high capacity stegnographic technique where secret data is embedded in intermediate significant bit planes besides least significant bit plane. The embedding scheme uses alternate top-down and down-top approach to improve the quality of stego-image. The security of the embedded data is taken care of by embedding data in all locations under the control of secret key. The experimental results show that proposed scheme performs better than some existing schemes.

Writer identification is the process of identifying the writer of the document based on their handwriting. Recent advances in computational engineering, artificial intelligence, data mining, image processing, pattern recognition and machine learning have shown that it is possible to automate writer identification. This paper proposes a model for text-dependent writer identification based on English handwriting. Features are extracted from scanned images of handwritten words and trained using pattern classification algorithm namely support vector machine. It is observed that accuracy of proposed writer identification model with Polynomial kernel show 94.27% accuracy.

In this paper, we study the variation of the convergence speed on basis of the local search method used in a memetic algorithm (MA); for this reason we compared two memetic algorithms MA1 and MA2 implemented with two different local search methods are descent and simulated annealing successively applied to the scheduling tasks problem. Then we will make a comparison between these two algorithms (MA1, MA2) results, with the genetic algorithm and other metaheuristics previously used to solve the same problem to get an idea on the most efficient methods in term of completion time work for this problem and choose the most effective.