Novel Algorithm for Hand Gesture Modeling Using Genetic Algorithm with Variable Length Chromosome

Many languages the people can exploit for them in order to communicate among them and get the message delivered, but, these languages should be known by those people in order to understand and speak, contrarily, gesture system is the common language that can be adopted for this objective and need less knowledge as compared with spoken languages that need the grammatically and semantically rules, in this paper we applied a novel algorithm for capturing hand gesture shape using one of the evolutionary algorithms in order to fit the hand segment. Previous techniques in the literature that fully captured hand shape applied some artificial intelligent methods [1] or some statistical methods [2]. Genetic Algorithms (GAs) with variable length of chromosomes is used to model the hand structure. The most effective GA parameters used for this purpose are; the generation of initial population, tournament selection, crossover with variable position of the cutting points in the parents, artificial mutation operator, deleting of the repetitive genes in same individual, and elitism strategy. Experimental results shows the robust and efficiency of applying the proposed algorithm. Keywordsgesture; feature extraction; genetic algorithms; initial population; variable chromosome length __________________________________________________*****_________________________________________________