西电高光谱机器学习大作业.docx

Hyperspectral remote sensing data classificationClass :Student ID:Name :AbstractCurrently, hyperspectral remote sensing (HRS) has become a cutting-edge field ofremote sensing (RS) area. Hyperspectral image classification has become one of thehottest research fields of the remote sensing area. The hyperspectral classificationtechnology has very high value both in application and theory research. However,hyperspectral image processing also introduced many problems. The problems are: howto solve the problem of high-dimensional with small amout of samples, how to use theinformation ignored in current research, such as structural information, how to use thevast amounts of unlabeled samples that contain information, how to choose the fittestclassifier. In HRS applications, the requirements of hyperspectral classifiersperformance are increasingly. The corresponding hyperspectral classificationtechnology research should focus on the topic, connecting with the specific applicationof hyperspectral remote sensing. IntroductionHyperspectral imaging spectrometer, also known as remote sensing, which uses a lot of very narrow band of electromagnetic waves reflected from the object of interest to obtain the relevant data to achieve a breakthrough in remote sensing technology to improve spectral resolution. Hyperspectral remote sensing and general difference is that the former than the latter has more finely spectral bands, you can form a complete spectral data. Technical staff through the spectral analysis of the data, the feature can be a lot of information, so as to achieve long-range, non-contact detection of objects conducted research. From the perspective of technological innovation, hyperspectral technology combines modern high-tech in a series of important progress, the use of the spectral characteristics of the surface features fine and the intrinsic link between the feature detection, resulting in remote sensing images each pixel corresponding to the spectral curv