Monocular Visual Place Recognition in LiDAR Maps via Cross-Modal State Space Model and Multi-View Matching