中文摘要：

设计了一种基于邻域激励脉冲耦合神经网络（PCNN）模型的图像分割方法．把既考虑强度又考虑邻域分布的像素邻域信息作为一个参数，来控制PCNN模型中的链接强度，进而控制神经元的内部活动值．在分割过程中采用基于多数裁定原则的方法，通过在一次迭代过程中对邻域像素分割阈值的调整，保证了分割结果的完整性．通过对几类图像的分割实验以及与经典分割方法的比较，验证了该方法的有效性．

英文摘要：

An approach to image segmentation was introduced by using neighborhood inspiring pulse coupled neural network （NIPCNN）. The neighborhoods considering brightness and distribution of the pixels were modeled into a factor to control the linking and the internal activity. Criteria based on majority rule controls the process, and threshold adjustment for heiborhoods within on iterative ensures the integrated result. Experiments of several types of images were implemented with the proposed method and the experimental results were compared with classical methods to demonstrate its validity.