arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1605.09136cs.CVcs.LGstat.ML
收藏

基于核分布嵌入的支持向量机高光谱图像分类

Hyperspectral Image Classification with Support Vector Machines on Kernel Distribution Embeddings

MINES ParisTech, PSL-Research University(巴黎矿业学院,PSL研究大学);University of Oxford(牛津大学)

Gianni Franchi, Jesus Angulo, Dino Sejdinovic

摘要:

我们提出了一种新的方法在高光谱图像中的像素分类,利用在数据中的空间和光谱信息。所介绍的方法依赖于最近提出的框架,用于学习分布-通过在再生核希尔伯特空间(RKHS)中用平均元素表示它们并在其中制定分类算法。特别是,我们将每个像素与其相邻像素的经验分布相关联,在RKHS中,结合像素本身所包含的光谱信息,给出了一组新的明确的特征,这些特征可以被馈送到一套标准分类技术中-我们选择了一个完善的支持向量机(SVM)框架。此外,通过随机傅立叶特征形式化降低了计算复杂性。我们研究了所提出的方法的一致性和收敛速度,实验表明,与最先进的结果相比,高光谱数据的增益具有很强的性能。

英文摘要:

We propose a novel approach for pixel classification in hyperspectral images, leveraging on both the spatial and spectral information in the data. The introduced method relies on a recently proposed framework for learning on distributions -- by representing them with mean elements in reproducing kernel Hilbert spaces (RKHS) and formulating a classification algorithm therein. In particular, we associate each pixel to an empirical distribution of its neighbouring pixels, a judicious representation of which in an RKHS, in conjunction with the spectral information contained in the pixel itself, give a new explicit set of features that can be fed into a suite of standard classification techniques -- we opt for a well-established framework of support vector machines (SVM). Furthermore, the computational complexity is reduced via random Fourier features formalism. We study the consistency and the convergence rates of the proposed method and the experiments demonstrate strong performance on hyperspectral data with gains in comparison to the state-of-the-art results.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑