COLOUR IS AN IMPORTANT VISUAL CUE TO UNDERSTAND SCENES. HOWEVER, MOST OF THE PREVIOUS RESEARCH IN COMPUTER VISION HAS BEEN MAINLY FOCUSED ON GREY LEVEL IMAGES. AT PRESENT, THIS TENDENCY HAS CHANGED AND COLOUR HAS BECOME AND ACTIVE RESEARCH TOPIC IN COMPUTER VISION. IN THIS WORK, WE FIRSTLY MAKE A REVIEW OF COLOUR PERCEPTION AND HOW TO DEAL WITH COMPUTATIONAL COLOUR. THE VARIATION OF THE ILLUMINATION CONDITIONS OF SCENES LEADS TO SERIOUS DIFFICULTIES WHEN WORKING WITH COLOUR IMAGES. WE HAVE CLASSIFIED THE DIFFERENT APPROACHES TO DEAL WITH THE VARIABILITY PROBLEM INTO THREE GROUPS: COLOUR CONSTANCY ALGORITHMS, DEFINITION OF COLOUR INVARIANTS AND COMPUTATION OF NORMALIZED IMAGES. SECONDLY, WE PRESENT A SIMPLE METHOD FOR COLOUR NAMING WHICH IS BASED ON A NORMALIZATION APPROACH, THAT IS THE COMPREHENSIVE COLOUR NORMALIZATION OF FINLAYSON ET AL., 1998). IN ORDER TO EVALUATE IT, OUR METHOD HAS BEEN TESTED ON A REAL SURVEILLANCE APPLICATION.COLOUR IS AN IMPORTANT VISUAL CUE TO UNDERSTAND SCENES. HOWEVER, MOST OF THE PREVIOUS RESEARCH IN COMPUTER VISION HAS BEEN MAINLY FOCUSED ON GREY LEVEL IMAGES. AT PRESENT, THIS TENDENCY HAS CHANGED AND COLOUR HAS BECOME AND ACTIVE RESEARCH TOPIC IN COMPUTER VISION. IN THIS WORK, WE FIRSTLY MAKE A REVIEW OF COLOUR PERCEPTION AND HOW TO DEAL WITH COMPUTATIONAL COLOUR. THE VARIATION OF THE ILLUMINATION CONDITIONS OF SCENES LEADS TO SERIOUS DIFFICULTIES WHEN WORKING WITH COLOUR IMAGES. WE HAVE CLASSIFIED THE DIFFERENT APPROACHES TO DEAL WITH THE VARIABILITY PROBLEM INTO THREE GROUPS: COLOUR CONSTANCY ALGORITHMS, DEFINITION OF COLOUR INVARIANTS AND COMPUTATION OF NORMALIZED IMAGES. SECONDLY, WE PRESENT A SIMPLE METHOD FOR COLOUR NAMING WHICH IS BASED ON A NORMALIZATION APPROACH, THAT IS THE COMPREHENSIVE COLOUR NORMALIZATION OF FINLAYSON ET AL., 1998). IN ORDER TO EVALUATE IT, OUR METHOD HAS BEEN TESTED ON A REAL SURVEILLANCE APPLICATION.