Collaborative information processing is the difficulty of the wireless sensor network research. To track a single target，a collaborative mechanism based on game theory is proposed，which creates a group way for the network nodes to execute target tracking，dynamically allocating tasks，updating the group members at any time，the synchronous information integration and data transfer. In the experiment, the collaborative mechanisms based on game theory and based on the instances are compared. Experimental results indicate that，target tracking based on game theory shows a significantly higher accuracy rate and significant reduction in network energy consumption; and its smooth transmission of information can adapt to the dynamic scalability and adaptability of the network nodes.