Abstract: To enhance the efficiency of Unmanned Aerial Vehicles (UAVs) in inspecting ship emissions, this study addresses the coordinated scheduling and routing planning problem for heterogeneous UAVs and Unmanned Surface vessels (USVs), employing a two-layer planning model to tackle the configuration and path optimization issues. The upper-layer model implements an improved adaptive simulated annealing algorithm to determine the allocation of UAVs and USVs. Meanwhile, the lower-layer model seeks to minimize the total flight time of UAVs by integrating the real-time dynamic characteristics of UAVs, USVs, and ships, utilizing an improved genetic algorithm to formulate the UAV detection sequence, flight path, and USV navigation path. Experimental results indicate that the coordinated scheduling of heterogeneous UAVs and USVs significantly reduces average flight time by 24.98%, UAV costs by 23.57%, and USV costs by 42.20% when compared to fixed configurations with homogeneous UAVs and USVs. Additionally, increasing UAV speed by 15 km/h results in an average reduction of 23.02% in flight time, while the UAV cost rises by 26.45%. Furthermore, for every 5 km/h increase in USV speed, the average USV cost escalates by 37%.