This talk presents an algorithm for combining multiple image segmentations to achieve a final segmentation. In contrast to previous works we consider the most general class of segmentation combination, i.e. each input segmentation has an arbitrary number of regions. Our approach is based on a random walker segmentation algorithm which is able to provide high-quality segmentation starting from manually specified seeds. We generate such seeds from an input segmentation ensemble. An information-theoretic optimality criterion is proposed to automatically determine the final number of regions. The experimental results on 300 images with manual ground truth segmentation will be demonstrated.