Figure 1
The three components of the video image analysis system. A: schematic representation; B and C: picture of the setup in our laboratory.
Figure 2
Observation and spectral analysis of a metronome at 200 oscillations per minute. A: observations from the first software module. B: mean power of observed frequencies; 3.35 Hz is the frequency that has the maximum mean power (200 opm = 3.333 Hz). C: spectrogram of observed frequencies over time (intensity scale on the right) (opm = oscillations per minute). D: correlation plot between theoretical and observed frequencies of a metronome (n = 24)
Figure 3
Spectral analysis of two periods of data acquisition with a metronome set up at 52 oscillations per minutes = 0.8667 Hz, in a continuous experiment. Left: spectrogram from data acquired at 9:00 hr. Right: spectrogram from data acquired at 18:50 hr.
Figure 4
Comparison of data from Gemvid and from Actiwatch. A: observed data from each device is presented in blue for Gemvid and red for Actiwatch. Mobile is going at 23.8 cm · s-1 and the mobile surface is 128 cm · s-1. B: correlation plot between the size of the moving surface and the mean number of pixels that changed for each value of area (n = 3 for each value of area).
Figure 5
Data from software module 1. A: screenshot showing the actual view of software module 1. Green pixels highlight zones that changed since the last frame. Movements highlighted here were only observed around the head and the tail. Left: the rat is active; right: the rat is sleeping. B: Examples of actograms. Top: one day actogram for rat #1 (LD 12:12 with lights on at 6:00). Below: seven days actogram for rat #3 (left) and #6 (right). C: mean number of pixels changed each day for one rat (left: in dark conditions ; right: in light conditions).
Figure 6
Activity of a rat in L:D and D:D conditions. Top: 3rd day in LD conditions ; bottom: 5th day under free-running, DD conditions ; a * denotes change in activity.