使用新型ICP演算法多視角重建實現機器人手臂的噴漆路徑

Abstract

這篇論文介紹了一種創新的三維模型噴塗方法，利用 Open3D 多視角重建來優化機械臂輔助的工業噴塗過程。該方法首先通過深度相機從多個角度捕獲詳細的三維點雲，並通過迭代最近點（ICP）算法進行精確對齊和重建。這種精細的對齊產生了高度準確的三維模型，能夠引導機械臂進行精確且高效的噴塗應用。此外，我們開發了一種基於生成的點雲模型的往復路徑規劃策略，模擬典型的噴塗運動模式，以提高效率、覆蓋均勻性和應用的一致性。通過實驗評估，我們證明了這種方法顯著提高了噴塗精度，確保均勻的覆蓋和提高過程的可靠性。這使得該方法非常適合於先進的自動噴塗應用，為滿足大規模生產環境中嚴格的工業質量和精度標準提供了一個可靠的解決方案。This paper introduces an innovative three-dimensional model painting method that utilizes Open3D multi-view reconstruction to optimize the spray painting process in industrial settings with the assistance of a robotic arm. The method initiates by capturing detailed 3D point clouds through depth cameras from multiple angles, which are then precisely aligned and reconstructed using the Iterative Closest Point (ICP) algorithm. This careful alignment results in highly accurate 3D models that guide the robotic arm for precise and efficient paint application. Additionally, we have developed a back-and-forth path planning strategy based on the generated point cloud model, simulating a typical painting motion pattern that enhances efficiency, coverage uniformity, and application consistency. Through experimental evaluation, we demonstrate that this approach significantly improves painting accuracy, ensuring even coverage and increased reliability across the process. This makes the method ideally suited for advanced automated spray painting applications, offering a robust solution to meet stringent industrial standards for quality and precision in large-scale production environments.