一种基于MAS的自适应图像分割方法

复杂图像的全自动分割是极具挑战性的问题．提出了一种基于MAS的自适应图像分割方法．通过属性和行为描述Agent个体，通过合作竞争描述了Agent间的交互．MAS系统在Agent反复自适应过程中达到平衡，同时完成图像分割．通过分割复杂的医学图像证实了该方法的有效性，MAS在图像分割领域具有应用价值．

An adaptive image segmentation technique based on multi-Agent system

The automatic segmentation of complex images is a challenging issue. A novel approach was proposed to segment images adaptively based on multi-Agent system. Individual agents were described with their attributes and behaviors. The interaction between Agents was described by cooperation and competi tion. Finally, the MAS reached a balanced state after a process of recursive self-adaptation and the image was segmented. Experiments with complex medical images show that the MAS approach is suitable for use in image segmentation.