Influenced by eutrophication and water system connectivity in typical lake-dominated water network areas (e.g., Taihu Lake), the invasive species Pomacea canaliculata frequently outbreaks in lakeside wetlands, park lakes, and rural water bodies. Due to its strong reproductive capacity and high environmental adaptability, Pomacea canaliculata has become widely established and continues to spread in this region, posing a serious threat to freshwater and agricultural ecosystems. Previous detection studies mainly relied on a single Unmanned Aerial Vehicle (UAV) or Unmanned Surface Vehicle (USV) platform, which struggled to balance large-scale screening and fine-grained identification due to limitations in viewing angle and operational efficiency. This study employs UAV–USV collaborative monitoring technology combined with deep learning methods to construct an “aerial screening–area projection–close-range verification” collaborative detection framework. Two independent datasets are established from UAV aerial imagery and USV close-range observations, covering typical water bodies, including lakeside zones, wetland parks, and rural villages. The results show that: (1) the proposed DBSCAN (Density-Based Spatial Clustering of Applications with Noise) density clustering method effectively generates community-level annotations, improving detection accuracy for dense small targets; (2) the improved UAV-side and USV-side models achieve mAP@0.5 values of 0.946 and 0.968, respectively, maintaining stable detection performance under complex water environments; (3) the air–water collaborative framework fully leverages the complementary advantages of UAV wide-area coverage and USV fine-scale observation, significantly improving detection efficiency and spatial localization accuracy. This study provides technical support for the precise prevention and control of Pomacea canaliculata in typical lake-dominated water network areas and offers a methodological reference for monitoring other aquatic invasive species in the future.