Image correspondence serves as the backbone for many tasks in robotics, such as visual fusion, localization, and mapping. However, existing correspondence methods do not scale to large multi-robot systems, and they struggle when image features are weak, ambiguous, or evolving. In response, we propose Natural Quick Response codes, or N-QR, which enables rapid and reliable correspondence between large-scale teams of heterogeneous robots. Our method works like a QR code, using keypoint-based alignment, rapid encoding, and error correction via ensembles of image patches of natural patterns. We deploy our algorithm in a production-scale robotic farm, where groups of growing plants must be matched across many robots. We demonstrate superior performance compared to several baselines, obtaining a retrieval accuracy of 88.2%. Our method generalizes to a farm with 100 robots, achieving a 12.5x reduction in bandwidth and a 20.5x speedup. We leverage our method to correspond 700k plants and confirm a link between a robotic seeding policy and germination.

翻译：图像对应是机器人视觉融合、定位与建图等任务的核心支撑。然而，现有对应方法难以扩展至大规模多机器人系统，且在图像特征微弱、模糊或动态变化时表现不佳。为此，我们提出自然快速响应码（Natural Quick Response Codes，简称N-QR），该方法能够实现大规模异构机器人团队间的快速可靠对应。N-QR的工作机制类似于QR码，通过基于关键点的对齐、快速编码以及自然图像块集成纠错来实现。我们在一个生产级机器人农场中部署了该算法，该场景下需要将多台机器人采集到的不同生长阶段的植物图像进行匹配。与多个基线方法相比，我们取得了88.2%的检索准确率，展现出优越性能。该方法可扩展至包含100台机器人的农场，带宽降低12.5倍，速度提升20.5倍。我们利用该方法实现了70万株植物对应，并证实了机器人播种策略与植物发芽率之间的关联性。