I 'd like to do band selection on hyperspectral images, i Found a lot of techniques in literature. which one is the most significant? and is there any libraries that can help me?
Perhaps my paper on the use of statistical information to do band selection using a technique I called the Optimum Index Factor (OIF) and linked below may be of help.
The selection of bands depend on your data analysis, if you are using it for regression modeling or supervised classification filter based approaches such as recursive feature elimination and genetic algorithm works very well.
I agree with Rajasheker's comments because band selection usually depends on your application. For your consideration, please find below three papers using three different approaches:
-Multispectral synthesis of daylight using a commercial digital CCD camera, J. L. Nieves, E.M. Valero, S.M.C. Nascimento, J.Hernández-Andrés and J. Romero, Applied Optics, Vol. 44, N. 27, pp. 5696-5703, (2005).
-Developing an optimum computer-designed multispectral system comprising a monochrome CCD camera and a liquid-crystal tunable filter, M.A. López-Álvarez, J. Hernández-Andrés, J. Romero, Applied Optics, vol. 47, 4381-4390 (2008).
-Evaluating a logarithmic kernel for spectral reflectance estimation - effects on model parametrization, training set size and number of sensor spectral channels, TT. Eckhard, E.M. Valero, J. Hernández-Andrés, V. Heikkinen, Journal of the Optical Society of America A, vol. 31, N.3, pp. 541-549 (2014).