Issue No. 10 - Oct. (2012 vol. 34)
Amir Alush , Bar-Ilan University, Ramt-Gan
Jacob Goldberger , Bar-Ilan University , Ramt-Gan
We present a method for combining several segmentations of an image into a single one that in some sense is the average segmentation in order to achieve a more reliable and accurate segmentation result. The goal is to find a point in the “space of segmentations” which is close to all the individual segmentations. We present an algorithm for segmentation averaging. The image is first oversegmented into superpixels. Next, each segmentation is projected onto the superpixel map. An instance of the EM algorithm combined with integer linear programming is applied on the set of binary merging decisions of neighboring superpixels to obtain the average segmentation. Apart from segmentation averaging, the algorithm also reports the reliability of each segmentation. The performance of the proposed algorithm is demonstrated on manually annotated images from the Berkeley segmentation data set and on the results of automatic segmentation algorithms.
Image segmentation, Clustering algorithms, Reliability, Correlation, Human factors, Optimization, Approximation algorithms, EM algorithm., Image segmentation, ensemble segmentation, integer linear programming, correlation clustering
J. Goldberger and A. Alush, "Ensemble Segmentation Using Efficient Integer Linear Programming," in IEEE Transactions on Pattern Analysis & Machine Intelligence, vol. 34, no. , pp. 1966-1977, 2012.