HLS Distorted colour model for enhanced colour image segmentation

In this work a new colour model HLS Distorted is presented which greatly improves the HLS colour model for use in robust colour image segmentation. When working with various colour models we often find that each has strengths and weaknesses for image analysis making most colour models useful only some of the time. The HLS colour model is one such model. While it is excellent at separating colour into components useful in many areas of computer vision it has the problem of being difficult to use when colours are weak. This weakness is caused by the colour modelpsilas axes converging at low saturation and low/high luminance causing calculated distances between perceptively similar colours to be larger than expected. For this reason the colour model is primarily used in applications working with brightly coloured objects. A technique is presented for distorting the HLS colour model in a unique way to keep the models strengths whilst removing its fundamental weakness.