论文部分内容阅读
利用基于视觉模型建立起来的局部激励、全局抑制振子神经网络 (OSNN)实现二值图象中数字及字母的目标 /背景分割 ;从理论上阐述了该网络实现目标 /背景分割的机理 ,并分析了如何合理选择参数以利于成功分割。
By using local excitation based on visual model, global oscillator and neural network (OSNN) can be used to realize digital / algebraic target / background segmentation in binary image. The mechanism of target / background segmentation in this network is theoretically analyzed and analyzed How to choose the parameters in order to facilitate the successful segmentation.