使用多個RGB-D攝像頭的3D重建系統，用於表型分析和植物運動監測

Abstract

這項研究引入了一個使用多個RGB-D攝像機來監測植物生長和運動的3D重建系統。該研究利用3D技術的潛力，精確捕捉植物對壓力的反應以及隨時間的生長模式。通過在植物周圍放置多個RGB-Depth攝像機，並結合捕獲的深度信息和RGB彩色圖像，研究創建了詳細的點雲，代表植物的結構。這些數據使植物在生長過程中能夠全面地建構數字模型，從而進行詳細觀察，如葉片運動。點雲註冊的一個關鍵方法是迭代最近點（ICP）方法。最終的目標是研究這種經濟實惠且易於獲取的3D重建方法，並評估其在溫室環境中的效率。此外，隨著在RGB-D攝像技術和圖像精度改進方面的進一步研究和進步，這個系統可以通過移動系統或靜態的更長範圍攝像機，改善監測技術和大規模的綜合建模，重建植物的數字複制品。

This research introduces a 3D reconstruction system using multiple RGB-D cameras to monitor plant growth and movements. The study utilizes the potential of 3D technology to capture precise data on plant responses to stresses and growth patterns over time. By employing multiple RGB-Depth cameras placed around the plant, and combining the captured depth information and RGB color images, the research creates detailed point clouds, representing the plant''s structure. This data allows for the comprehensive construction of a digital model of the plant throughout its growth, allowing for detailed observations such as leaf movements. A critical method in the registration of these point clouds is the Iterative Closest Point (ICP) method. The ultimate aim is to study this affordable and accessible 3D reconstruction method and evaluate its efficiency in environments like greenhouses. Moreover, with further research and advancements in RGB-D camera technology and image accuracy improvements, this system could improve monitoring techniques and comprehensive modelling on a larger scale through a moving system or static longer-range cameras, recreating digital replicas of the plants.