基于演化算法和超限学习机的高光谱遥感图像信息处理研究

针对高光谱遥感图像中存在的维度灾难和非线性数据等复杂问题的挑战，本项目拟开展基于演化算法和超限学习机的结合算法研究，深入探讨二者之间存在的相互作用与影响，并将研究成果应用于高光谱遥感图像信息处理的实际问题中。在全面分析演化算法和超限学习机在高光谱遥感图像应用研究现状的基础上，创新性地开展以下主要研究内容：1）针对高光谱遥感图像波段选择问题，开展基于超限学习机的演化算法模型研究；2）针对高光谱遥感图像特征提取问题，开展基于演化算法的深度超限学习机模型研究；3）针对高光谱遥感图像分类识别问题，开展基于演化算法的新型超限学习机模型研究；4）自主设计并开发基于演化算法和超限学习机的高光谱遥感图像信息处理原型系统，并将前述算法集成于该系统。本项目的研究为高光谱遥感图像信息处理提供有效的解决方案，同时有利于进一步推进演化计算、机器学习和高光谱遥感图像应用等领域的交叉研究，具有重要的理论意义和应用价值。

For hyperspectral remote sensing images, they suffer from the challenges of the curses of dimensionality and the nonlinear data structure. Motivated by these, in this proposal we will investigate the integrated algorithms, which are the combination of evolutionary algorithm and Extreme Learning Machine for the applications of hyperspectral remote sensing image, to explore the underlying relationship and interaction between evolutionary algorithm and Extreme Learning Machine. Through analysis of the current research status and the pros and cons of evolutionary algorithm and Extreme Learning Machine in the applications of hyperspectral remote sensing image, we mainly focus on the follow key issues: (1) Evolutionary algorithm based on Extreme Learning Machine for the problem of band selection in hyperspectral remote sensing image; (2) Deep extreme learning machine based on evolutionary algorithm for the problem of feature extraction in hyperspectral remote sensing image; (3) Extreme learning machine based on evolutionary algorithm for the problem of classification in hyperspectral remote sensing image; and (4) The software about information processing for hyperspectral remote sensing image. Therefore, we can believe that our research in this proposal is very important and meaningful, and it will promote the cross-research in evolutionary computation, machine learning and the applications of hyperspectral remote sensing image.