Colour image segmentation is an important tool in many applications, like robotics, computer vision and data compression. Differing highly from grey-scale images, the colour segmentation usually has more complicated and time consuming algorithms and is controlled by a larger set of parameters. The measures of quality of colour segmentation are presented in the article, enabling users to evaluate the most efficient set of parameters for a procedure of colour segmentation.The evaluation of parameters are suggested to be provided by heuristical and statistical methods, also presenting relationship of such measures.