Pentagram Research Center

Hyderabad, India

Pentagram Research Center

Hyderabad, India

Time filter

Source Type

Rao Y.N.,Anna University | Vemuganti Y.,Staffordshire University | Rajan E.G.,Pentagram Research Center
Proceedings: Turing 100 - International Conference on Computing Sciences, ICCS 2012 | Year: 2012

This paper deals with the formulation of a non-Chosky grammar called M-grammar which defines the language of End Justified Post Turing Rewriting Systems. Formulation of this grammar has solved the Problem of Generating Markov Class Rewriting Systems By a Type-0 Phrase Structure Grammar. © 2012 IEEE.


Dimlo U.M.F.,Anna University | Vemuganti Y.,Staffordshire University | Rajan E.G.,Pentagram Research Center
Proceedings: Turing 100 - International Conference on Computing Sciences, ICCS 2012 | Year: 2012

This paper introduces novel concepts like rewriting cyclic normal automata and flower automata. The principle of normalization of automata states that for any given automaton belonging to a family of automata one can construct a equivalent rewriting cyclic normal automaton. © 2012 IEEE.


Inturu L.P.,Pentagram Research Center | Prashanthi G.,Staffordshire University | Rajan E.G.,Pentagram Research Center
Proceedings: Turing 100 - International Conference on Computing Sciences, ICCS 2012 | Year: 2012

In this paper, we demonstrate the use of the string manipulating techniques in non-numerical signal processing. First, we construct a normal algorithm for carrying out cyclic shifting in a string of arbitrary length. Next, we outline a general method for constructing normal algorithmic systems for signal processing operations on symbolic sequences in general. As an illustration, we then construct a normal algorithmic system for carrying out linear convolution of nonnegative integer sequences. © 2012 IEEE.


Ashok J.,Gurunanak Institute of Technology | Rajan E.G.,Pentagram Research Center
European Journal of Scientific Research | Year: 2012

Handwriting has continued to persist as a means of communication and recording information in day-to-day life even with the introduction of new technologies. This paper describes performance comparison of Rajan transform and Radial basis function on offline hand written character recognition. Importance of handwriting in human transactions, machine recognition of handwriting has practical significance, as in reading handwritten notes in a personal Digital Assistant (PDA), in postal addresses on envelopes, in amounts in bank checks, in handwritten fields, in forms etc. Algorithms of preprocessing, character and word recognition, and performance with practical system are indicated. The comparision has been made among Radial Basis Function (RBF) and Rajan Transform and found to be better compared to that of rate in the proposed system lies between 98% to 100%. © EuroJournals Publishing, Inc. 2011.


Rao S.S.,Auroras Technological and Research Institute | Rajan E.G.,Pentagram Research Center | Lalkishore K.,Jawaharlal Nehru Technological University
Wireless Personal Communications | Year: 2012

RFID tags are used for varied applications in large numbers. Human Tracking is one such important application wherein the RFID system detects the presence of a tag in a dense tag environment. Quick estimation of the number of tags in the field at a desired level of accuracy is one of themost common requirements in the present scenario. Identification of tags has become a critical area which need more time and unsuitable inmany situations; especially where tag set is dense. We introduce a novel medium access control (MAC) protocol for Radio Frequency IDentification (RFID) systems which exploits the statistical information collected at the reader. The protocol, termed Adaptive Slot Adaptive Frame (ASAF) ALOHA protocol, is motivated to significantly improve the total read time performance of the currently suggested MAC protocols for RFID systems by minimizing the collisions of the tags. In order to perform this task, ASAF estimates the dynamic tag population and adapts the frame size and number of slots simultaneously in the subsequent round via a simple policy that maximizes an appropriately defined function.We demonstrated that ASAF is better than the current RFID MAC protocols.We also considered the case where suddenly if the number of tags increases, the running frame with adapted slots gets flushed away and new frame with increased number of slots gets adapted automatically by the system showing the robustness in this case as well. © Springer Science+Business Media, LLC. 2011.


Tallapragada V.V.S.,C.B.I.T. | Rajan E.G.,Pentagram Research Center
Proceedings - 1st International Conference on Integrated Intelligent Computing, ICIIC 2010 | Year: 2010

Iris biometric is considered to be one of the most efficient and trusted biometric method for authenticating users. This paper presents a method for iris recognition based on hybrid feature set of wavelet and GLCM features, a non-filter based technique, combined with Haar wavelet transform to increase the efficiency of the system. Here we combine frequency domain feature with spatial domain feature to increase overall efficiency of system. Probability neural Network is used to classify the features. Results show that the overall system efficiency is 94% with false rejection rate higher than false acceptance rate. The technique is tested on CASIA Iris database. © 2010 IEEE.


Tallapragada V.V.S.,C.B.I.T. | Rajan E.G.,Pentagram Research Center
ACE 2010 - 2010 International Conference on Advances in Computer Engineering | Year: 2010

Network security and protection of data have been of great concern and a subject of research over the years. There are many different forms of cryptographic mechanisms like AES, Triple DES, MD5 proposed to guarantee data security. In a network, the success of the algorithm depends on the length of the key that user uses. It is observed that due to convenience of remembering the key, users use short keys. This increases the vulnerability of the data. In this work we propose a unique authentication and encryption technique using IRIS biometric pattern of a person. At the time of registration to the network, a person's IRIS is scanned and phase features of the image are generated using Log-Gabor filter. Filter output is 64 × 64 vectors which is difficult to use as key or to transmit over the network for authentication. Therefore a local binary code of sixty four byte is extracted of this matrix and is transmitted to the server. As this is the basis of entire algorithm functioning, we use a secured socket layer with SH1 and 128 keys(Browser supported) for this communication. Server stores this key as users Identity or password. When he wants to transfer data to another user, first his IRIS pattern is generated, it is sent to the server, server authenticates the user by comparing the incoming IRIS code with the one saved in database by using Euclidian distance. Once authenticated, user is allowed to exchange information over the network. Every information is now be encrypted using the receivers IRIS code with MD5. The success of the technique depends upon the Quality of the scan and recognition rate. There is always an error margin of 5% to 10% when IRIS code is generated which may result in an unsuccessful authentication. Therefore at the time of registration, we scan the person's eyes thrice and two codes are embedded in a smart card as reference code. This smart card is issued to the user by the administrator. At the time of authentication, user code is first compared with the code of smart card and if matched (Even with an error margin of 10%) then only his code is transmitted to server for authentication. Storing the rough key in the server is not safe as if the server data is hacked, then these keys can be easily tampered with. Hence rather than storing the rough keys, we embed them in random images using steganographic means and store the image binary in the server. Thus not only the data in the network, but also the server data is secured. © 2010 IEEE.


Subramanyam T.V.,Pentagram Research Center | Selvarajan S.,Muthayammal Technical Campus | Rajan E.G.,Pentagram Research Center
Proceedings: Turing 100 - International Conference on Computing Sciences, ICCS 2012 | Year: 2012

Mathematical objects of analysis are of two types, numbers and words. Number based calculations are essentially arithmetic operations of addition, subtraction, multiplication and division. Symbol or words based calculations are symbol manipulating algorithms. This paper presents certain elementary manipulating techniques using which one can construct algorithms. © 2012 IEEE.


Satyanarayana Tallapragada V.V.,C.B.I.T. | Rajan E.G.,Pentagram Research Center
2015 International Conference on Pervasive Computing: Advance Communication Technology and Application for Society, ICPC 2015 | Year: 2015

With the technological advancement security lapse is of major concern. Hence different techniques are adopted to provide better security. In this juncture, biometrics is widely used. Iris is one of such biometric which can provide high security when compared to other existing biometric traits. In this paper we propose a novel segmentation method for segmenting the iris part which is occluded and can be seen partially. Proposed segmentation has resulted in 90% accurate segmentation over MMU Iris database and with 1.8 seconds time for segmenting each iris. Further different features are extracted from the segmented iris part and are combined to form a feature vector. These are classified using decision tree classifier. Results show improved performance when compared to the existing techniques. © 2015 IEEE.


Tallapragada V.V.S.,C.B.I.T. | Rajan E.G.,Pentagram Research Center
IET Image Processing | Year: 2012

IRIS biometric is one of the most efficient and trusted biometric methods for authenticating users owing to invariance with age or with physical activities. IRIS recognition techniques are broadly categorised in three groups: phase, texture and kernel-based methods, which out of kernel-based methods are proven to be the best suited for IRIS recognition problem. In this work a multiclass kernel Fisher analysis and its consequent feature set for IRIS recognition is proposed. The authors use support vector machine (SVM) classifier to group the large database into smaller groups where each group is linearly separable from the other. Once an image is grouped as one of the groups by SVM, it is classified to be recognised by hidden Markov model (HMM) classifier which compares the features of the given image only with the other images of the same group. Results show 93.2% overall accuracy for the system if we consider seven features and improved to 99.6% when 1200 features are used. In order to meet this efficiency an average convergence time needed by the algorithm is found to be lesser than existing SVM-based technique. Results also show fast convergence time for optimisation process in comparison to with other conventional kernel and SVM-based techniques. © 2012 The Institution of Engineering and Technology.

Loading Pentagram Research Center collaborators
Loading Pentagram Research Center collaborators