In this paper we propose a three-stage algorithm to implement effective segmentation of an object’s counterparts when illumination variance is present. A color constancy algorithm and color-based invariant parameters insensitive to a large set of different illumination conditions are used. Then reflectance images are considered after discarding the shadowing information present in the images. A color-based segmentation algorithm using Graph Cuts is applied to the reflectance images. Improvements in the segmentation are obtained after using these illumination invariants.
CITATION STYLE
Pérez-Carrasco, J. A., Acha-Piñero, B., Serrano-Gotarredona, C., & Gevers, T. (2014). Reflectance-based segmentation using photometric and illumination invariants. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 8814, pp. 179–186). Springer Verlag. https://doi.org/10.1007/978-3-319-11758-4_20
Mendeley helps you to discover research relevant for your work.