论文部分内容阅读
最佳指数法是常用的高光谱图像数据波段选择方法,但存在运算时间过长的问题。运用K-means聚类算法,对最佳指数方法进行了改进,提出了聚类最佳指数法,并进行了一系列伪装目标识别的对比实验。实验结果表明,与最佳指数法相比,改进后的方法在保证目标分类精度的前提下,运算速度提高了数十倍;与单纯使用K-means聚类运算相比,不仅运算时间缩短,而且分类精度有所提高。利用改进算法能够在伪装环境下更加快速有效地识别目标。
The best index method is a commonly used method of band selection of hyperspectral image data, but there is a problem that the operation time is too long. Using K-means clustering algorithm, the best index method is improved, the best clustering index method is proposed, and a series of comparative experiments of camouflage target recognition are carried out. Experimental results show that compared with the best index method, the improved method can improve the computational speed by tens of times under the premise of ensuring the accuracy of target classification. Compared with simply using K-means clustering algorithm, not only the computing time is shortened, Classification accuracy has improved. Improved algorithms are used to identify targets more quickly and effectively in camouflage environments.