In order to address the challenges of environmental noise, complex distribution of water surface targets, and the blurring of small-scale features in water surface target detection under complex river backgrounds, this paper proposes an enhanced water surface target detection algorithm called UltraWS, which integrates multi-scale features and attention mechanisms. Firstly, a spatial attention module and multi-head strategy are designed on a typical detection network to fuse multi-scale features and improve the detection capability of small targets. Secondly, the UltraLU module is introduced to enhance class activation mapping and reduce the influence of environmental and distribution factors on target detection. Finally, a Tucker tensor decomposition method is applied to achieve model lightweighting, enhancing model interpretability and inference speed. Experimental results demonstrate that the proposed UltraWS algorithm improves the resistance to background noise, captures small targets better, and achieves a balance between detection speed and accuracy for edge deployment requirements. On the WSODD dataset, the algorithm achieves the highest mAP value of 84.5%, outperforming other mainstream methods by an considerable improvement. The proposed algorithm, along with the established channel safety inspection system and evaluation method, contributes to the development of intelligent river transportation.