重新审视2D-3D配准：一种用于高价值区域选择与表征一致性对齐的新网络
Rethinking 2D-3D Registration: A Novel Network for High-Value Zone Selection and Representation Consistency Alignment
用强化学习解决高价值匹配区域选择不可微的问题，并用三阶段（理解-协调-加速）做图像纹理与点云结构间的区域表征对齐，可直接迁移到跨模态配准与稠密对应任务中抑制非重叠区域误匹配。
原文依据 HZRS employs reinforcement learning to resolve the non-differentiable issue of selecting high-value matching regions
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论先检测后匹配与免检测两类方法在图像到点云配准中均已被广泛研究，但仍面临显著挑战。先检测后匹配的方法强调高质量对应关系，却受限于可重复关键点的可得性，容易受到错误匹配带来的误差影响。相比之下，免检测方法借助由粗到细的策略追求稠密对应以缓解匹配误差。然而，非重叠区域与低质量匹配仍会引入不准确，且图像纹理与点云结构之间的差异导致区域表征不一致，从而提高了错误匹配的似然。为应对这些挑战，我们提出两个创新模块：高价值区域强化选择模块（HZRS）与区域表征一致性对齐模块（ZRCA）。HZRS 采用强化学习来解决高价值匹配区域选择不可微的问题，ZRCA 则通过理解、协调与加速三个阶段改进区域对齐。在 RGB-D Scenes v2 与 7-Scenes 上的大量实验与消融研究证明了我们网络的优越性，使其成为图像到点云配准的最优方法。
英文原文摘要 1325 字符
Both detection-then-match and detection-free methods have been extensively studied for image-to-point cloud registration, yet they still face significant challenges. The detection-then-match approach emphasizes high-quality correspondences but is limited by the availability of repeatable keypoints, making it susceptible to errors from incorrect matches. In contrast, detection-free methods aim for dense correspondences using a coarse-to-fine strategy to mitigate matching errors. However, non-overlapping regions and low-quality matches still introduce inaccuracies, and the differences between image texture and point cloud structure cause inconsistent region representations, increasing the likelihood of incorrect matches.To address these challenges, we propose two innovative modules: the High-Value Zone Reinforced Selection Module (HZRS) and the Zone Representation Consistency Alignment Module (ZRCA). HZRS employs reinforcement learning to resolve the non-differentiable issue of selecting high-value matching regions, while ZRCA improves region alignment through three stages: understand, coordinate, and accelerate.Extensive experiments and ablation studies on RGB-D Scenes v2 and 7-Scenes demonstrate the superiority of our network, establishing it as the state-of-the-art for image-to-point cloud registration.