Handwritten Mathematical Symbol Recognition Based on Niche Genetic Algorithm

Abstract

Handwritten character recognition is a kind of technology that recognizes the handwritten symbols automatically. Aim at the complexity and limitations of the traditional handwritten character recognition method, a NGA-BP model for character recognition which combines the niche genetic algorithm and neural network is built. This method makes great use of the searching ability of ecological niche genetic algorithm and the nonlinear mapping and associative ability of BP neural network, it extracts the coarse grid characteristics, the projector features, cross-cut characteristics and structural features, then makes use of the operation of choice, cross, variation and obsolete of the ecological niche genetic algorithm, optimizes the initial weight values and threshold of BP neural network, finally, makes the well-trained NGA-BP network recognize the mathematical symbols. The test results show that the recognition method of combining the ecological niche genetic algorithm and the artificial neural network in this article is a good solution to the problem irrational network initial value, and it symbolizes the complexity of algorithm, increases the speed of network convergence. Compare with traditional ways, this model has great feasibility and validity, also has higher recognition rate and better reliability.