关键词: 噪声受益, 新型期望最大化算法（NEM）定理, 图像分割, 空间邻域关系, 改进的高斯混合模型

Abstract: Image segmentation denotes a process by which a raw image is partitioned into nonoverlapping regions. When using the existing improved Gaussian mixture model in image segmentation, how to speed up its segmentation process is a significant research topic. Based on the latest noise-benefit EM algorithm, this paper speeds up the convergence speed of the existing improved Gaussian mixture model by adding artificial noise, which achieves the goal of speeding up image segmentation. Additive noise speeds up the average convergence of the EM algorithm to a local maximum of the likelihood surface when adding noise to meet the noise-benefit EM theorem. Improved Gaussian mixture model is a special case of the expectation-maximization algorithm, therefore, noise-benefit EM theorem applies to improved gaussian mixture model. Experimental results indicate that the algorithm speeds up the convergence speed when it is used for image segmentation, and the time complexity is decreased significantly.

Key words: noise benefit, New Expectation Maximization（NEM） theorem, image segmentation, spatial neighborhood relationships, improved Gaussian mixture model