In this paper, we treat one of the central problems in computer vision and pattern recognition witch is the image segmentation. We adapt the Multi-objective Bacterial Foraging Optimization Algorithm (MBFOA) to optimize simultaneously two thresholding criteria, Between-Class Variance 2D and Entropy 2D, and to improve the quality of the segmentation. In this paper, we treat one of the central problems in computer vision and pattern recognition witch is the image segmentation. We adapt the Multi-objective Bacterial Foraging Optimization Algorithm (MBFOA) to optimize simultaneously two thresholding criteria, Between-Class Variance 2D and Entropy 2D, and to improve the quality of the segmentation.