Stitching of hyper-spectral UAV images based on feature bands selection

This study employed the feature band pre-selection and composite process to overcome the mosaic difficulties of large number of bands and huge data volume of the hyperspectral UAV images before the mosaic process being done. The application and comparison were done by using the UHD 185 hyperspectral camera, and the result showed that the band pre-selection and composite method used in this study can go up as high as being 390 percent faster than the single band mosaic (for the Normalized Differential Vegetation Index application), and about 46 times faster than full band mosaic. Besides, compared to the full band mosaic method, the method used in the study can reduce the maximum memory usage amount and final mosaic file size significantly.