Fuzzy Logic and Object Identification of Wharf Yard Construction Mechanical Injury Risk Monitoring and Warning

The construction environment of wharf yards is highly complex, with significant risks of mechanical injuries, underscoring the need for effective monitoring and early warning systems. Existing methods, such as the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) and You Only Look Once version 3 (YOLOv3), often suffer from limited accuracy and poor real-time performance, making them unsuitable for dynamic conditions. To address this gap, this study proposes a risk monitoring and early warning approach that integrates fuzzy inference with target recognition. Fuzzy inference was used to quantify and classify mechanical injury risks, while the You Only Look Once version 5 (YOLOv5) detection algorithm and Simple Online and Realtime Tracking with a Deep Association Metric (Deep SORT) technique enabled accurate recognition and tracking of workers and machinery. By combining fuzzy logic–based risk levels with target recognition outputs, the system delivers real-time early warnings and countermeasure recommendations, providing a robust solution for enhancing human–machine safety in dynamic construction environments.