Colour Similarity Measure at Different Brightness Using CIELUV Colour Space

Abstract

<p>In the field of colour science and image processing, quantifying colours into meaningful information can be hard because the perception of colour is an individual experience and it is difficult to measure. This paper proposed a method of comparing colours perceptually based on the CIELUV colour space. The Colour Similarity Measure (CSM) was created after observing the movement of the point of the same colour in the CIELUV space at different external brightness. This algorithm can judge and map the similarity of colour between two images while minimizing the lighting effect. With a simple modification, the CSM can also be used as an edge detection algorithm and the proposed method is proven to have a human-like sensibility of edge in an image. Even without the help of deep learning, we demonstrate the ability of CSM to find edge patterns almost as good as the current best edge detector in the colour blind test. </p>