【摘要】 提出了如何用一种有生物视觉依据的人工神经网络———脉冲耦合神经网络(PulseCoupledNeuralNetwork,PCNN)进行二值文字空洞滤波,即将二值文字中的空洞填满,从而便于后续的文字识别。计算机仿真结果表明,利用PCNN的脉冲传播特性,可快速而有效地将二值文字中的空洞填满,故基于PCNN的空洞滤波是一种有效的空洞滤波新方法。更多还原

【Abstract】 This paper brings forward how to do binary character hole-filter, namely fill holes of binary characters by using PCNN(Pulse Coupled Neural Network), which has the biologically visual background,so that the next character recognition can be realized.The results of computer simulations show that by using PCNN,holes of binary characters can be filled efficiently and quickly.Therefore,Hole-filter based on PCNN is a new efficient approach for binary character hole-filter.更多还原