While deep learning-based image matching methods excel at extracting high-level semantic features from remote sensing data, their performance degrades significantly under cross-daylight conditions and wide-baseline geometric distortions, particularly in multi-source street-view scenarios. This paper presents a novel illumination-invariant framework that synergistically integrates geometric topology and semantic consistency to achieve robust multi-view matching for cross-daylight urban perception. We first design a self-supervised learning paradigm to extract illumination-agnostic features by jointly optimizing local descriptors and global geometric structures across multi-view images. To address extreme perspective variations, a homography-aware transformation module is introduced to stabilize feature representation under large viewpoint changes. Leveraging a graph neural network with hierarchical attention mechanisms, our method dynamically aggregates contextual information from both local keypoints and semantic topology graphs, enabling precise matching in occluded regions and repetitive-textured urban scenes. A dual-branch learning strategy further refines similarity metrics through supervised patch alignment and unsupervised spatial consistency constraints derived from Delaunay triangulation. Finally, a topology-guided multi-plane expansion mechanism propagates initial matches by exploiting the inherent structural regularity of street scenes, effectively suppressing mismatches while expanding coverage. Extensive experiments demonstrate that our framework outperforms state-of-the-art methods, achieving a 6.4% improvement in matching accuracy and a 30.5% reduction in mismatches under cross-daylight conditions. These advancements establish a new benchmark for reliable multi-source image retrieval and localization in dynamic urban environments, with direct applications in autonomous driving systems and large-scale 3D city reconstruction.

Cross-modal semantic segmentation of individual tree LiDAR point clouds is critical for accurately characterizing tree attributes, quantifying ecological interactions, and estimating carbon storage. However, in forest environments, this task faces key challenges such as high annotation costs and poor cross-domain generalization. To address these issues, this study proposes a cross-modal semantic transfer framework tailored for individual tree point cloud segmentation in forested scenes. Leveraging co-registered UAV-acquired RGB imagery and LiDAR data, we construct a technical pipeline of “2D semantic inference—3D spatial mapping—cross-modal fusion” to enable annotation-free semantic parsing of 3D individual trees. Specifically, we first introduce a novel Multi-Source Feature Fusion Network (MSFFNet) to achieve accurate instance-level segmentation of individual trees in the 2D image domain. Subsequently, we develop a hierarchical two-stage registration strategy to effectively align dense matched point clouds (MPC) generated from UAV imagery with LiDAR point clouds. On this basis, we propose a probabilistic cross-modal semantic transfer model that builds a semantic probability field through multi-view projection and the expectation–maximization algorithm. By integrating geometric features and semantic confidence, the model establishes semantic correspondences between 2D pixels and 3D points, thereby achieving spatially consistent semantic label mapping. This facilitates the transfer of semantic annotations from the 2D image domain to the 3D point cloud domain. The proposed method is evaluated on two forest datasets. The results demonstrate that the proposed individual tree instance segmentation approach achieves the highest performance, with an IoU of 87.60%, compared to state-of-the-art methods such as Mask R-CNN, SOLOV2, and Mask2Former. Furthermore, the cross-modal semantic label transfer framework significantly outperforms existing mainstream methods in individual tree point cloud semantic segmentation across complex forest scenarios.

Accurate semantic segmentation of LiDAR point clouds from urban river levees is critical for flood hazard assessment and infrastructure maintenance, yet dense vegetation cover severely occludes the terrain and introduces ambiguous return signatures, challenging conventional filtering and learning-based methods. To address this, we propose TCSF-Net, a hybrid framework that synergistically integrates a Transformer backbone with a cloth simulation filtering (CSF) prior. The Transformer captures long-range contextual dependencies across irregular point clouds, while the CSF module embeds physical constraints that mimic cloth draping over the surface, guiding the network to preserve terrain continuity and suppress vegetation outliers, especially under canopy occlusion. Experimental results demonstrate that TCSF-Net consistently outperforms state-of-the-art methods, including pure Transformer architectures and classical CSF, in overall accuracy, mean intersection-over-union, and boundary preservation, with marked improvements in sloped transition zones and near-water edges. Our results further show that the fusion of data-driven learning with physically grounded regularization effectively mitigates the ambiguity caused by multi-layer vegetation. We conclude that integrating physical simulation priors into deep segmentation networks offers a robust and generalizable solution for high-fidelity terrain extraction and vegetation discrimination in complex fluvial environments.