In the field of resource-constrained robots and the need for effective place
recognition in multi-robotic systems, this article introduces RecNet, a novel
approach that concurrently addresses both challenges. The core of RecNet's
methodology involves a transformative process: it projects 3D point clouds into
depth images, compresses them using an encoder-decoder framework, and
subsequently reconstructs the range image, seamlessly restoring the original
point cloud. Additionally, RecNet utilizes the latent vector extracted from
this process for efficient place recognition tasks. This unique approach not
only achieves comparable place recognition results but also maintains a compact
representation, suitable for seamless sharing among robots to reconstruct their
collective maps. The evaluation of RecNet encompasses an array of metrics,
including place recognition performance, structural similarity of the
reconstructed point clouds, and the bandwidth transmission advantages, derived
from sharing only the latent vectors. This reconstructed map paves a
groundbreaking way for exploring its usability in navigation, localization,
map-merging, and other relevant missions. Our proposed approach is rigorously
assessed using both a publicly available dataset and field experiments,
confirming its efficacy and potential for real-world applications.

该研究介绍了一种名为 RecNet 的新方法，通过将三维点云投影到深度图像并使用编码器 - 解码器框架进行压缩，从而同时解决资源受限机器人和多机器人系统中有效的位置识别问题。通过提取该过程中的潜在向量用于高效的位置识别任务，RecNet 不仅取得了可比较的位置识别结果，而且保持了紧凑的表示形式，适合在机器人之间实现无缝共享以重建其集体地图。通过对 RecNet 进行包括位置识别性能、重建点云的结构相似性和仅共享潜在向量而得到的带宽传输优势在内的一系列指标进行评估，该方法为在导航、定位、地图合并和其他相关任务中的应用开辟了新的途径。通过公开数据集和实地实验对所提出的方法进行了严格评估，证实其在真实世界应用中的有效性和潜力。