The pantograph slide plate (PSP) plays a crucial role in the daily maintenance of trains by transmitting electrical energy from the contact wire (CW) to the train. Automatic monitoring of PSP wear condition is essential to notice abnormal wear promptly. In field applications, the target PSP edge may be obscured by the rear side PSP due to the operating attitude of the pantograph, leading to false detections and interfering with the field maintenance work. Meanwhile, the scenario of the PSP occlusion is characterized by randomness, which brings a considerable challenge to the reconstruction of the occluded edges. A robust PSP wear monitoring method is proposed to improve the performance of wear detection in field applications, considering the PSP occlusion scenarios. Firstly, the wear area of the PSP is localized using the YOLOv5s model. Subsequently, a fusion of the morphology method and the Canny operator is employed to extract the edge of the PSP. Different occlusion types are identified through detailed observation, and corresponding edge reconstruction methods are proposed for occluded edges. Finally, a robust wear condition evaluation method and five evaluation criteria for assessing wear condition are introduced. Experimental results demonstrate that the system achieves accurate PSP wear detection with an accuracy of 0.6 mm, and the reconstructed edge at occlusion locations closely matches the target edge compared to other automatic systems.