Movement characterization with the spatiotemporal Fourier transform of low-light-level images.

The spatiotemporal Fourier transform of a series of frames is proved to be an attractive technique for object velocity determination. The main disadvantage of the spatiotemporal transform methods is that they require a long processing time even when fast transforms are used. We show that the spatiotemporal Fourier-transform technique, which has been applied to standard-light-intensity images, can also be used with low-light-level images. The result is an important processing time reduction; thus when fast processing is required, it would be advantageous to detect in low-light-level conditions with intensity filters. The technique behavior for low-light-level images is checked with computer-simulated and experimental data.