论文部分内容阅读
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邻域的局部信息，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。
Image segmentation and object extraction are an important part of image processing system. In this paper, a target extraction algorithm based on Kohonen self-organizing neural network is proposed. It not only makes use of the overall gray-scale information of the image, but also uses the local information of each pixel neighborhood. The method is implemented on the 486 microcomputer using C language and can extract the target under noisy conditions.