一种改进型脉冲耦合神经网络及其图像分割1

本文结合人类视觉系统(HVS)对图像信息含量区域敏感度不同这一特性,以神经元接近点火程度的一致性描述图像空间邻域所含的信息量,对通常的脉冲耦合神经网络模型(PCNN－Pulse Coupled Neural Network)进行了改进,提出了一种基于改进PCNN的图像自适应分割算法.该算法根据象素及其周边区域的信息量大小发放不同值的脉冲,从而自适应地将图像分为多个不同等级的高低信息区域,较好地仿真了人类视觉系统特性.最后对用这种方法进行图像分割的结果进行基于信息量的图像压缩,在压缩比和重建图像主观视觉感知质量上均达到了良好的性能,表明了本文算法的可行性和有效性.

Image Segmentation Based on a Modified Pulse-Coupled Neural Network

Based on the property of Human Vision System(HVS) that human eye's sensitivity to an image varies with different informative regions of the image,Pulse Coupled Neural Network(PCNN) model is modified for image segmentation.The modified PCNN stimulated by an input image has pulse output with multiple pulse values rather than two,according to the local information rate of the input image.This results in segmentation of the image with respect to local information rate delivered by the image.Experiments on image segmentation with our method and image compression on the segmented images that the lower the information rate is,the higher the compression rate is,show much better performance in both compression rate and subjective perceptual quality of the reconstructed image compared to image segmentation with standard PCNN and image compression with an overall compression rate.