| 改进的SGA端元选择的快速方法 |
| |
| 引用本文： | 王立国,邓禄群,张晶.改进的SGA端元选择的快速方法[J].应用科技,2010,37(4):19-22. |
| |
| 作者姓名： | 王立国 邓禄群 张晶 |
| |
| 作者单位： | 哈尔滨工程大学,信息与通信工程学院,黑龙江,哈尔滨,150001 |
| |
| 基金项目： | 国家自然科学基金，教育部博士点新教师基金资助项目，水下智能机器人技术国防科技重点实验室联合资助项目 |
| |
| 摘 要： | SGA算法因其自动性和高效性受到广泛欢迎,针对该算法包含大量的体积计算导致该算法的运算速度较慢的问题,采用在高维空间中构造超平面的方法,提出了一种SGA的改进方法.该改进算法把复杂的体积比较转化成简单的点到超平面的距离比较,从而将算法的复杂度由空间维度的3次关系降至线性关系.实验表明,快速SGA与原始SGA在端元选择的结果上保持一致,而在端元选择的速度上前者较后者有大幅度提高,尤其是要选择的端元数目越大时,效果越明显.
|
| 关 键 词： | 高光谱图像 端元选择 SGA 距离比较 |
A fast endmember selection method based on simplex growing algorithm |
| |
| WANG Li-guo,DENG Lu-qun,ZHANG Jing.A fast endmember selection method based on simplex growing algorithm[J].Applied Science and Technology,2010,37(4):19-22. |
| |
| Authors: | WANG Li-guo DENG Lu-qun ZHANG Jing |
| |
| Institution: | WANG Li-guo,DENG Lu-qun,ZHANG Jing(College of Information , Communication Engineering,Harbin engineering University,Harbin 150001,China.) |
| |
| Abstract: | SGA has been a popular one for its full automation and high-efficiency performance.An improved SGA method was proposed and a method of constructing a hyperplane was adopted in the high-dimensional space,for the purpose of improving a low speed algorithm which involves innumerable volume calculation and thus leads to low speed of calculation.Complex volume comparison was converted to simple comparison of the distance from a point to a hyperplane in this improved method.As a result,the complexity of fast SGA ... |
| |
| Keywords: | SGA |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
