Geometric Correction Algorithm for Hyperspectral Images based on GPS

A geometric correction algorithm based on GPS for the hyperspectral images acquired by push-broom imaging spectrometer is proposed. The correction method consists of three steps, namely, data import, coordinate computing and image correction. Data import read hyperspectral data and attitude information data given by GPS; coordinate computing calculates the Gaussian plane rectangular coordinates for each pixel in the image; while image correction calculates the corresponding coordinate in the corrected image of the pixels according to the Gaussian plane rectangular coordinates and using the nearest neighbor interpolation method to smooth the image. The corrected results show that the algorithm is ideal for hyperspectral image geometric correction, and the algorithm is computational efficiency.