SEVENT FRAMEWORK
PROGRAMME
MT’ M
M = D65 s T D65
RGB RGB XYZ
However, once more, the temperature transform MT’ to take us from D65 RGB to T RGB is
not easy to find directly, and as strictly speaking, matrix multiplication is non-commutative,
we Cannot simply invert this as M’ + M * MT’. However, if M is not too far from an identity
matrix and T is not far from D65 then: M’ = M * MT, where MT->MT’ and yields an
approximate M’.
= x
RGB [ds
1 PANE
fy RGB
Calibration matrix M converts
camera’s RGB output illuminated M x M 5
at a colour temperature T into D65
CIE XYZ A colour temperature adjustment
input image into the calibration
matrix M is D65 and not Tas in
VIS image
Figure 1-23. Schematic showing the colour correction of visible images (left) and the application of the
calibration matrix M together with a colour temperature transform, MT (to compensate for the
difference in colour temperature between D65 and T) for the correction of UV-induced visible
luminescence images (right).
| is required to compensate that
Version No. 1.0 29 Date : 14/10/2013