To address the requirements for detection accuracy and real-time performance in identifying small-sized broken wires and wear defects on the surface of steel wire ropes, the algorithm is optimized based on the YOLOv9 framework. By introducing the PGI auxiliary supervision mechanism to enhance gradient information propagation, constructing a GELAN for efficient feature extraction, and designing a Multi-Scale Fusion Module along with a Dynamic Sparse Attention Mechanism, the model's feature representation capability for tiny and occluded targets is significantly improved. Experimental validation was conducted using the Version dataset, and the results demonstrate that the improved model achieves a detection accuracy of 90.2% and an inference speed f 587.43, representing a significant performance enhancement over the original YOLOv9. The algorithm maintains excellent detection performance for surface defects on steel wire ropes, providing a reliable technical solution for intelligent detection of broken wires and wear on steel wire rope surfaces.