基于ANN端元估计的高光谱图像解混算法

Unmixing of hyperspectral images based on endmember estimation of artificial neural network

针对高光谱图像解混问题进行研究,发现传统解混算法在保持端元数目不变的情况下,得到的解混精度不高。为此,基于人工神经网络(artificial neural network,ANN)提出一种估计单像素点中端元数目和类别的解混算法。首先利用人工神经网络对遥感图像中各个像素的端元数目和类别进行估计;之后依据估计结果确定解混算法的目标函数,并引入改进的差分搜索算法对目标函数进行优化求解;最终获取地物丰度和待求参数,实现高光谱图像的解混。仿真数据和真实遥感数据实验表明,与现有的解混算法相比,所提解混算法具有更高的解混性能,更加符合实际场景的情况。

Aimed at the problems of hyperspectral images unmixing,it is found that the unmixing accuracy of the traditional unmixing algorithm is not high when the number of endmember keeps constant in unmixing.Thus,based on the ANN,this paper proposed a novel unmixing algorithm of estimating the number and category of endmember in a single pixel.Firstly,the unmixing algorithm used the ANN to estimate the number and category of each mixed pixel’s endmember in the remote sensing image.Then,it determined the objective function of the algorithm based on the estimation results,and introduced the improved differential search algorithm to solve the objective function.Finally,it obtained the abundances and the parameters to realize the unmixing of hyperspectral images.The experimental results on simulated and real hyperspectral data demonstrate that compared with the existing unmixing algorithms,the proposed unmixing algorithm has higher performance and is more in line with the actual scene.