Background subtraction with multispectral video sequences

Motion analysis of moving targets is an important issue in several applications such as video surveillance or robotics. Background subtraction is one of the simplest and widely used techniques for moving target detection in video sequences. In this paper, we investigate the advantages of using a multispectral video acquisition system of more than three bands for background subtraction over the use of trichromatic or monochromatic video sequences. To this end, we have established a dataset of multispectral videos with a manual annotation of moving objects. To the best of our knowledge, this is the first publicly available dataset of multispectral video sequences. Experimental results indicate that using more than three spectral sub-bands provide better discrimination between foreground and background pixels. In particular, the use of the near infra-red (NIR) spectral band together with visible spectra provides the best results.