Influence of noise, light and shadows on image segmentation algorithms
José Cecílio, Pedro Furtado, Pedro Martins · International Journal of Applied Pattern Recognition · 2019
Image segmentation is a required step in object recognition tasks, dividing the image into multiple regions or clusters. Each image pixel is assigned to one of the clusters using different metrics such as pixel colour value, grey-scale intensity, edges, shapes, among others. The most diverse image segmentation algorithms have been proposed. It is important to analyse the robustness of those algorithms, taking into consideration diverse types of noise and difficult conditions. In this paper we choose a small set of frequently-used algorithms and analyse their behaviour in robustness-testing conditions. We inject difficulties (noise, shadows, various degrees of illumination) and compare the quality of the segmentation of algorithms against a ground truth. The objective is to analyse how differences in illumination, shadows or noise influence the output of the algorithms, and how they compare on those metrics. Based on those results we conclude about the quality of the approaches tested.