Results and Discussion
300 400 X (pixels)
Figure 8.8 The positions in the camera’s FOV of marker are shown with error bars representing the spread in X and Y magnified by 1000. It appears that the markers near the edge of the FOV have slightly higher standard deviations. This is consistent with the results in Figures 8.6a, b.
σY(X) = 46.5 × 10
31.3 × 10
6X + 15.9 × 10
The data were also analysed in terms of distance from the centre of the FOV. Despite the fact that the centre of the FOV is not precisely known, it appears that a similar trend exists as shown for a varying X value but in this case it is for a varying radial distance.
Variation due to distance
The Black Spot modules are designed to track markers in their FOV at a distance of between 0.5 m and 3 m. Therefore, it is important to test the modules’ performance between these distances. Black Spot module is fixed at distances of 0.5, 1, 1.5, 2, 2.5, and 3 m with respect to the test beacon, facing the markers directly. The standard deviations of the centroids over 2 minute periods were calculated at the different distances. Table 8.1 shows the results obtained. The mean of the centroid standard deviation of each marker is measured. This is given along with a measure of spread from the mean as an indicator of how reliable the
s shown in Table 8.1, the spreads in standard deviation for the 1 m and 0.5 m measure- ments are significantly higher than the rest. The data for these distances were analysed and it was determined that one marker was producing much higher centroid variations than other markers at this distance. This marker had a low maximum intensity (significantly