A Two-Level Method for Combining Segmentation Maps Based on Information Redundancy Measure
Dmitry M. Murashov · 2022 VIII International Conference on Information Technology and Nanotechnology (ITNT) · 2022
In this paper, we propose a new two-level method for combining image segmentation maps based on the minimum information redundancy criterion. The method makes it possible to obtain a more accurate partition of the original image and preserve informationally important areas that may be lost during the operation of traditional segmentation algorithms. The results of the experiment showed that the proposed method improves the result of image segmentation in terms of minimizing information redundancy and informational dissimilarity with the original image and ground-truth segmentation maps.