earticle

영어

Hyperspectral Image processing is usually time consuming, due to its huge data size. Nowadays Hyperspectral Imaging is used in many fields where real-time solutions are required. A systemic comparison study of linear feature extraction methods for classification of hyperspectral images with various types of noises is carried out in this paper, in which the performance of different linear feature extraction methods for classification and their computation cost reduction are compared. In practice, hyperspectral images are often contaminated by different types of noises, as the atmosphere around hyperspectral cameras may change all the time. In this paper, to make it more realistic, different types of noises, including Salt-and-Pepper noise, Gaussian noise, Speckle noise and their mixtures, are artificially imposed on the hyperspectral image. Support Vector Machine based classification is employed for classification performance comparison. The experimental results are very helpful for selecting linear feature extraction methods for classification of hyperspectral images that are usually affected with noises.