摘要图像分割是图像识别和计算机视觉的基础，得到精确的分割效果至关重要，因此，如何准确分割目标区域成为数字图像处理研究的难点和热点。图像分割算法很多，其中基于脉冲耦合神经网络的图像分割是基于哺乳动物的视觉皮层同步脉冲发放现象而构建的模型，依赖于图像的自然特性，得到广泛的研究和发展。40925
本文简单介绍传统PCNN的数学模型，通过仿真实验说明图像分割过程，分割效果表明传统PCNN难以得到理想的分割效果。为了解决这一问题，从生物角度考虑，本文重新定义动态阈值，构建新颖的PCNN的数学模型，其主要思想为：当动态阈值缓慢下降时，所有神经元从初始的全局抑制缓慢向点火状态转化，且某一神经元一旦点火，则以后步数都处于点火状态。
为了说明改进方法的有效性，与传统PCNN模型进行仿真对比实验，仿真结果说明改进算法的分割效果优于传统PCNN算法。
毕业论文关键词:    图像分割    动态阈值    脉冲耦合神经网络
Abstract
Image segmentation is the base of image recognition and computer vision, get accurate segmentation result is important, therefore, how to accurately segment the target area become a difficulty and hotspot in the research of the digital image processing. Many image segmentation algorithm, which based on pulse coupled neural network image segmentation is based on the mammalian visual cortex sync pulse distribution phenomenon and building model, depends on the nature of the image and get a wide range of research and development.
The paper simply introduces the mathematical model of the traditional PCNN, through the simulation experiment shows that image segmentation process, segmentation results show that the traditional PCNN is hard to get ideal segmentation results. In order to solve this problem, from a biological point of view, this paper redefines the dynamic threshold, build a mathematical model of the novel PCNN, its main idea is: when the dynamic threshold slow decline, all neurons to slowly from the initial global inhibiting ignition state transformation, and once a certain neurons ignition, steps, after all in a state of ignition.
To illustrate the effectiveness of the improved method, simulation experiments with the traditional PCNN model, the simulation results show that the improved algorithm is superior to the traditional PCNN algorithm of segmentation effect.
Keywords:   Image segmentation   Dynamic threshold
            Pulse coupled neural network  
目录
摘要    I
Abstract    II
1 绪  论    1
1.1课题研究的背景和意义    1
1.2图像分割的研究现状    2
1.3 PCNN的国内外研究进展    2
2 脉冲耦合神经网络基本理论    5
2.1 PCNN理论    5
2.2 PCNN基本模型    6
2.2.1 生物视觉模型    6
2.2.2 脉冲耦合神经元模型    8
3 基于PCNN的图像分割    11
3.1 图像分割的定义    11
3.2 图像分割方法    12
3.2.1 阈值分割方法    12
3.2.2 基于边缘的分割方法    13
3.2.3 基于区域的分割方法    13
3.3 基于PCNN图像分割算法实现    14
4 PCNN及其图像分割    17
4.1 改进的PCNN模型    17
4.2 点火模式    18
4.3 模型的行为分析    19
4.4 分割终止准则    21
4.5实验过程    22 基于PCNN脉冲耦合神经网络的图像分割:http://www.751com.cn/tongxin/lunwen_40790.html