| 迷彩设计中基于SOFM的背景图像聚类方法 |
| |
| 引用本文： | 初苗,田少辉,喻钧,陈宏书,胡志毅.迷彩设计中基于SOFM的背景图像聚类方法[J].计算机应用与软件,2009,26(10):218-221. |
| |
| 作者姓名： | 初苗 田少辉 喻钧 陈宏书 胡志毅 |
| |
| 作者单位： | 1. 西安工业大学,陕西,西安,710032
2. 总后勤部建筑工程研究所,陕西,西安,710032 |
| |
| 摘 要： | 迷彩设计要求用简单的几种颜色模拟自然背景纹理,因此,通过分析背景的图像,利用图像分割技术提取背景信息从而进行迷彩图案的自动设计已是近年来伪装领域研究的热点课题.目前,最常用的图像分割方法是均值聚类法,而均值聚类法由于直接利用图像的灰度差信息,容易造成分割的细节不明显,从而导致模拟背景纹理失真.针对均值聚类法的不足,结合人工智能网络学习技术,提出基于SOFM的图像分割方法对背景进行聚类分割,较好地保留了图像的细节纹理,得到了较为理想的效果.
|
| 关 键 词： | 迷彩设计 均值聚类 自组织特征映射网络 |
SOFM BASED BACKGROUND IMAGE CLUSTERING IN CAMOUFLAGE PATTERN DESIGN |
| |
| Chu Miao,Tian Shaohui,Yu Jun,Chen Hongshu,Hu Zhiyi.SOFM BASED BACKGROUND IMAGE CLUSTERING IN CAMOUFLAGE PATTERN DESIGN[J].Computer Applications and Software,2009,26(10):218-221. |
| |
| Authors: | Chu Miao Tian Shaohui Yu Jun Chen Hongshu Hu Zhiyi |
| |
| Affiliation: | Xi'an Technological University;Xi'an 710032;Shaanxi;China;Architectural Engineering Institute of the General Logistics Department;China |
| |
| Abstract: | Camouflage pattern design requires using a simple pattern with several colours to imitate nature background textures,so automatic design of camouflage pattern has been the hotspot research subject in camouflage field in recent years which abstracts the background information using image segmentation technique through analysing image of the background.Currently the most common image segmentation method is the mean clustering,however it is easily to cause the segmented details vague due to directly utilising ... |
| |
| Keywords: | |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
