Geodesic Distance Field-Based Five-Axis Continuous Sweep Scanning Method for the Multi-Entrance Inwall Surface

Abstract

Multi-entrance Inwall (MEI) surfaces are widely used in industrial applications, yet inspecting the MEI surfaces precisely remains a challenging task due to their complex multi-entrance topology and potential collision risks. The recently developed five-axis continuous sweep scanning technology offers significantly higher inspection efficiency compared to traditional point-by-point methods, presenting a valuable opportunity for accurate and efficient MEI surface inspection. However, planning a five-axis continuous sweep scanning process for general MEI surfaces still largely relies on human input. To address this challenge, this paper presents a novel set of methods for generating five-axis sweep scanning paths, specifically designed for the automatic and efficient inspection of MEI surfaces. Our methodology utilizes a sophisticated heat-induced geodesic distance field (GDF) to calculate guiding curves, which are used to generate the sweep scanning paths and partition the accessible regions of the MEI surface. This approach results in the creation of continuous five-axis sweep scan inspection paths that enhance both inspection efficiency and surface coverage rates. The proposed method has been validated through physical inspection experiments and computer simulations, with results confirming its feasibility and demonstrating its advantages over two benchmark approaches. Note to Practitioners—This article aims to generate an automatic, highly efficient inspection path for MEI surfaces using a five-axis coordinate measuring machine (CMM). While existing methods have addressed some issues in free-form surface inspection, they primarily focus on external and open surface inspection and do not adequately adapt to MEI surfaces, which are often occluded by challenging collision situations and complex topology. As a result, planning a five-axis continuous sweep scanning process for a general MEI surface still heavily relies on human interaction. To address this limitation, we propose an inspection path generation method that constructs a set of guiding curves considering the geometric information of both the entrances and collision situations. We utilize a heat-induced Geodesic Distance Field (GDF) to compute guide paths and assist in partitioning the accessible region. Through experiments and computer simulations, our proposed method demonstrates superior performance compared to traditional benchmarking methods in terms of both inspection efficiency and point accessibility rate. The generated inspection path conforms to the workpiece surface geometry, effectively overcoming challenges such as high interference and topological complexity in MEI, thus enabling efficient and comprehensive surface measurements.