用无需选取参数的Unit-linking PCNN进行自动图像分割

脉冲耦合神经网络（PCNN-Pulse Coupled Neural Network）是一种有生物学依据的人工神经网络，它可有效地用于图像分割。基于PCNN的图像分割效果取决于PCNN中各参数的选择。然而，图像分割时，各种不同的图像对应的PCNN参数是不同的，而PCNN参数的选择是困难的。本文提出了一种基于Unit-linking PCNN的图像分割新方法，解决了PCNN图像分割参数选择的难题。用本文提出的新方法可有效地自动分割各种图像，而无需考虑PCNN参数的选择，这对于PCNN的理论研究和实际应用有重要的意义。

Automatic image segmentation using Unit-linking PCNN without choosing parameters

PCNN-pulse coupled neural network, a novel artificial neural network supported by biology, can be efficiently applied in image segmentation. The performance of image segmentation based on PCNN depends on the suitable PCNN parameters. However, it is difficult to get the suitable PCNN image segmentation parameters because different kinds of images have different suitable PCNN parameters. In this paper, we propose a novel approach for image segmentation based on Unit-linking PCNN, by which we can use the same PCNN parameters to efficiently segment different kinds of images. Therefore, using this approach can segment images without choosing different parameters for different kinds of images automatically and efficiently.