In response to the decreased accuracy in person detection caused by densely populated areas and mutual occlusions in public spaces, a human head-detection approach is employed to assist in detecting individuals. To address key issues in dense scenes—such as poor feature extraction, rough label assignment, and inefficient pooling—we improved the YOLOv7 network in three aspects: adding attention mechanisms, enhancing the receptive field, and applying multi-scale feature fusion. First, a large amount of surveillance video data from crowded public spaces was collected to compile a head-detection dataset. Then, based on YOLOv7, the network was optimized as follows: (1) a CBAM attention module was added to the neck section; (2) a Gaussian receptive field-based label-assignment strategy was implemented at the junction between the original feature-fusion module and the detection head; (3) the SPPFCSPC module was used to replace the multi-space pyramid pooling. By seamlessly uniting CBAM, RFLAGauss, and SPPFCSPC, we establish a novel collaborative optimization framework. Finally, experimental comparisons revealed that the improved model’s accuracy increased from 92.4% to 94.4%; recall improved from 90.5% to 93.9%; and inference speed increased from 87.2 frames per second to 94.2 frames per second. Compared with single-stage object-detection models such as YOLOv7 and YOLOv8, the model demonstrated superior accuracy and inference speed. Its inference speed also significantly outperforms that of Faster R-CNN, Mask R-CNN, DINOv2, and RT-DETRv2, markedly enhancing both small-object (head) detection performance and efficiency.

This study investigates the spatiotemporal evolution and key obstacle factors of urban flood resilience in the Shenyang Metropolitan Area, aiming to inform regional flood resilience planning and management. A comprehensive assessment indicator system was established, integrating natural, economic, social, and infrastructure dimensions to capture the multifaceted nature of flood resilience. The long short-term memory (LSTM) network with an attention mechanism, combined with the obstacle degree model, was employed to analyze resilience trends and diagnose limiting factors from 2001 to 2023. The findings reveal a sustained increase in the regional flood resilience index, rising from 0.255 in 2001 to 0.574 in 2023. Spatially, the resilience pattern evolved from a monocentric core diffusion to a dual-core leadership and multi-city collaborative structure, driven by basin-wide management and differentiated development between mountainous and plain areas. Disparities in resilience levels across cities narrowed over time. At the criterion level, infrastructure was the primary obstacle before 2010, while social factors became increasingly significant thereafter. At the indicator level, the main limiting factors varied among cities and shifted over time, reflecting local development dynamics. These results provide a theoretical basis and practical guidance for enhancing urban flood resilience in the Shenyang Metropolitan Area and offer insights applicable to other rapidly urbanizing regions.