Figure 8.

The figure's top row shows reconstructed frames from events acquired by each sensor of the three-chip event-based camera. The gray levels are tone-mapped with a logarithmic function in order to be displayed on a regular screen. The bottom-left frames are reconstructed with the linear color model presented in the methodology. The colors are properly reconstructed, but very little detail is left in dark areas. The bottom-right frames use the channels' logarithmic tone-mappings as their color channel, yielding incorrect colors but much more details in dark areas.