| 基于Kohonen神经网络的目标提取 |
| |
| 引用本文： | 陈丽亚,戚飞虎.基于Kohonen神经网络的目标提取[J].上海交通大学学报,1995,29(6):24-28. |
| |
| 作者姓名： | 陈丽亚 戚飞虎 |
| |
| 作者单位： | “区域光纤通信网与新型光通信系统”国家重点实验室光纤技术研究所 |
| |
| 摘 要： | 图像分割与目标提取是图像处理系统的一个重要组成部分，本文提出了一种基于Ｋｏｈｏｎｅｎ自组织神经元网络的目标提取算法，它不仅利用了图像的整体灰度分布信息，而且利用了每个像素领域的局部信息，该方法在４８６微机上用Ｃ语言进行了实现，在有噪声的情况下也能提取目标。
|
| 关 键 词： | 神经网络 图像分割 目标提取 |
Object Extraction Using Kohonen Neural Network |
| |
| Chen Liya,Qi Feihu.Object Extraction Using Kohonen Neural Network[J].Journal of Shanghai Jiaotong University,1995,29(6):24-28. |
| |
| Authors: | Chen Liya Qi Feihu |
| |
| Institution: | Chen Liya;Qi Feihu |
| |
| Abstract: | Segmentation and object extraction are important parts of the image processing system. In this paper we present an object extraction algorithm by using using Kohonen neural network. In the algorithm proposed here, the grey level information of each pixel together with its neighborhood is selected as the input of the Kohonen neural netWork, which works without a supervisor. The network converges after weight adjusting and the pikels are classified into two groups: object and background. Experimental results have shown that it works well. |
| |
| Keywords: | Kohonen neural network image segmentation object extraction |
| 本文献已被 CNKI 维普 等数据库收录！ |
