To meet the requirements of many industries for not only the profile information but also the colour information of subjects, a system of colouring is established. Errors caused by the difference of the geometrical properties of the measured surface in the process of assigning colour directly are researched. In order to account for the change in the geometrical properties of the measured surface, it is divided into several sub-units, so the change of the geometrical properties of the measured surface can be considered as the change in the relative position among the sub-units and the projector, as well as the camera. The mathematical model is then established based on the positional relation. The relation between the slope change of the sub-units and their incident intensity and scattered intensity is analyzed. Finally, by combining with the impact of the relation on the colour information, the selection principle for the sub-units is determined to eliminate the errors of the colouring brought by the impact of the incident and the scattered intensities. Further, a colouring procedure that is not affected by the incident and scattered intensities is given. The simulation and experimental application of the colouring are executed by making use of the 3D measured data given by the 3D measurement system. The results show that the incident and scattered intensities of the different parts of the measured surface is corrected uniformly, and the errors in R, G, B colour weights are less than 0.02 (the values of R, G, B range between 0 and 1). Colouring information of the measured surface is almost uniform compared with its real colour, and even by changing the measuring angle, there is still a good visual effect.