HomeContact Us

Accepted Papers


  • Spanish Scrabble Findings
    Alejandro Gonzalez Romero1, Rene Alquezar1, Arturo Ramirez Flores2, Francisco Gonzalez Acuna2 and Ian Garcia Olmedo3, 1Universitat Politecnica de Catalunya, Barcelona, Spain, 2Centro de Investigacion en Matematicas A.C, Guanajuato, Mexico, 3Independent Researcher
    The game of Scrabble has been tackled by computers mainly by using simulation [1], [2] and [3]. In [4] an alternative method which uses a heuristic function that involves probability calculations to evaluate moves is presented. Following these ideas a heuristic algorithm was developed to make Heuri [] (a computer program that plays Scrabble). This paper presents useful results for Spanish Scrabble obtained recently by obtaining statistics of the data gathered when playing thousands of Heuri vs Heuri games. The data gathered by these games along with appropriate stats on them, help in the understanding of the nature of the game. The results also give useful information to humans which helps them in becoming stronger Scrabble players. It could even help a certain human to achieve world championship.
  • Online Signature Recognition Using Neural Network
    Babita Pegu and Aditya Bihar Kandali, Dibrugarh University, India
    Here it discusses about some of the features of signature data and their extraction from the raw data set collected from ATVS signature database. The features include time duration, sign changes of dx/dt and dy/dt, average jerk, number of pen-up pen-down etc. Signature features are pre-processed and brought to a value having same decimal point and trained using back propagation neural network. For signature data of 10 users and accuracy rate of 86% is obtained.
  • Spatial Filtering and Morphological Operation as Pre-Processing Steps in Fingerprint Feature Extraction
    Himangkana Goswami and Aditya Bihar Kandali, Jorhat Engineering College, India
    Extracting features from a fingerprint image relies mainly on the pre-processing stages the fingerprint has gone through. When the fingerprint image that has been captured is good enough then the final matching stage will produce a satisfying output. But many a times the image which is captured suffers from contact problems such as non-uniform contact, inconsistent contact and irreproducible contact.Because of such adverse and unpredictable image acquisition situations, a biometric system’s (Fingerprint Recognition System) performance suffers from random false rejects/accepts. Hence the need for the pre-processing of an image becomes necessary. In this paper, pre-processing steps of spatial filtering and morphological operation in addition to Gabor filtering are introduced and comparative analyses of the three are done in MATLAB. It has been found that there is a significant removal of false minutiae in the step of minutiae extraction, if spatial or morphological filtering methods are introduced prior to Gabor filtering.
  • Optimization of a Function in Genetic Algorithm: Survival of the Fittest
    Shabia Shabir Khan, S.M.K.Quadri and M.A.Peer, Kashmir University, India
    In the field of optimization, Genetic Algoithm that incorporates the process of evolution, plays an important role in finding the best solution to a problem.This paper discusses various processes behind genetic algorithm, the parameters and options used. Different computational experiments have been conducted to check the performance of the genetic algorithm by changing the values of parameters. It shows how the little variations in the values for particular parameter changes the objective function vaue for the genetic algorithm.
  • Fusion of ANFIS with Genetic Algorithm for Optimization
    Shabia Shabir Khan, S.M.K.Quadri and M.A.Peer, Kashmir University, India
    Due to the complexities in data, extracting knowledge out of it, intelligently, is one of the greatest challenges in the field of data mining. Several Artificial intelligent techniques like fuzzy logic or neural network have been provided to deal with such problems. However such techniques work properly and the desired result is met under certain conditions. As far as Fuzzy Logic is concerned the major issue lies in building up of membership functions and making decision on its appropriate parameters (MFs, their distribution and composition of fuzzy rules) and that in Neural Network, the architecture is an important issue. So taking such issues into consideration, the hybrid system of the two is used wherein GA optimization technique is also used that optimizes various parameters like learning rate, momentum, number of membership functions for each input. These soft computing techniques when combined prove to be faster than the individual ones.
  • Embedding and Extraction Techniques for Medical Images- Issues and Challenges
    S.Priya1 and R.Varatharajan2,1Bharath University, India, 2Sri Lakshmi Ammal Engineering College, India
    The new technologies in multimedia and communication fields have introduced new ways to transfer and save the medical image data through open networks which has introduced new risks of inappropriate use of medical information. Medical images are highly sensitive hence secured transmission and reception of data is needed with minimal distortion. Medical image security plays an important role in the field of Telemedicine. Telemedicine has numerous applications in teleconsulting, teleradiology, telediagnosis, telesurgery and remote medical education. Our work is to analyze about the different embedding techniques that can be used for embedding the personal and diagnosed details of a person within the medical images without any virtual discrepancy. Also to survey about the blind extraction algorithm utilizing genetic algorithm for optimization of the key parameters.
Copyright ® AIAA 2015