HSI–MSER: Hyperspectral Image Registration Algorithm based on MSER and SIFT

Requirements

• Ubuntu 18.04 or later

Dependencies

• Install: sudo apt-get install libmatio-dev

Usage

Run ./HSIMSER with the following parameters:

./HSIMSER <reference_image.mat> <target_image.mat> <number_of_bands_to_select>

<number_of_bands_to_select> is the number of bands to select in the first stage of the algorithm. In the article, 8 is chosen as the default value. Other figures can be used.

Output:

• Textual result in console: scale, angle of rotation, and shifts over x and y axis.

Additional program

rot_scal allows generating new target images with a specific scale and rotation angle to test the performance of the method.

Use:

./rot_scal <image.mat> <scale> <angle>

The output image is saved in the same directory with the name rot_scal_out.mat.

Examples

Using Santa Barbara Line images:

./HSIMSER SantaBarbaraLine2013.mat SantaBarbaraLine2014.mat 8

Using Pavia University image:

1. ./rot_scal PaviaUniversity.mat 2.0 90
2. ./HSIMSER PaviaUniversity.mat ./rot_scal_out.mat 8

Citation

This algorithm was presented in the article “HSI–MSER: Hyperspectral Image Registration Algorithm based on MSER and SIFT”, which is under review in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing. If you use this program in your research projects, we encourage you to please cite our work:

@article{alvaro2021hsimser,
    title = {{HSI-MSER: Hyperspectral Image Registration Algorithm based on MSER and SIFT}},
    journal = {IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing},
    author = {Ord{\'o}{\~n}ez, {\'A}lvaro and Acci{\'o}n, {\'A}lvaro and Arg{\"u}ello, Francisco and Heras, Dora B}
}