Speedup of fuzzy Co-clustering algorithm for image segmentation on graphic processing unit

The fuzzy co-clustering algorithms are used to solve the problem clustering of large data, multi-dimension, multifeature. When the dimension and size of data increases, the size of the membership function matrix increases, the number of calculations increases, computational problems become a burden for the computer. In fuzzy co-clustering algorithm, multiple rules and sub-algorithms may be executed in parallel by the GPU to speedup for the whole system. This paper studies the structure of fuzzy co-clustering algorithm and proposed a new solution how to implement fuzzy co-clustering algorithms on GPU. GPU-based calculations are solutions to improve computational efficiency and independent operation for the CPU. The study is demonstrated through experiments on color image clustering on CPU and GPU. The experiments were shown that implementing fuzzy co-clustering on GPU is much faster than on CPU. © 2015 ACM.