The algorithm for dynamic target allocation of cross-domain unmanned swarm in an imprecise information environment is proposed in this paper. Firstly,The paper delineats a realistic scenario of cross-domain drone swarm warfare and conducting an analysis of the inherent uncertainties in target positioning and impact points of firepower units due to the imprecision of detection information. Secondly, a probability model for target pre-allocation is developed based on this analysis, followed by the design of an enhanced discrete multi-objective particle swarm algorithm for solution purposes. Additionally, to address real-time emergence of new targets in the combat environment, a contract net-based target re-allocation algorithm, employing market mechanisms, is proposed to enable the real-time updating of target allocation schemes. Finally, the efficacy of the proposed algorithms is validated through experimental simulations. This research offers a comprehensive solution for the dynamic target allocation challenge encountered in cross-domain unmanned swarm operating in an imprecise information environment, with the potential to enhance the effectiveness and efficiency of unmanned swarm warfare.