Abstract: This paper introduces P system for image thresholding segmentation and proposes a thresholding segmentation approach based on P system. It uses a two-tier tissue-like P system as its computing framework and every membrane is regarded as an independent computing unit, and uses the speed-location model as its evolution rule. At the same time, the communication rule with symport/antiport rule is employed to exchange and share the objects among the membranes. The framework of the P system can find optimal segmentation thresholding for images. Experimental results illustrate the applicability and effectiveness of the proposed thresholding segmentation method.