Application and Evaluation of Precision in Food Ink Pattern Printing Utilizing Image-Guided Non-Planar Slicing Technology

Abstract

The implementation of robotic arms utilizing non-planar slicing technology for the extrusion and printing of starch-based ink streamlines the workflow for operators without specialized expertise in slicing methodologies. However, the model design stage often requires manual positioning of pattern points, which can result in similarity issues in the printed products, particularly for irregular patterns. Addressing this challenge necessitates designing printing patterns without requiring extensive professional skills. To overcome this obstacle, our approach uses image calibration to assist in pattern design. Accurate trajectory mapping between the intended design and the actual pattern was achieved by transforming the image coordinate system into the manipulator coordinate system. This “hand-eye collaboration” in pattern design and printing offers a novel solution for food additive manufacturing patterns and shaping design. The visual equipment was developed, and its calibration accuracy was optimized. On this basis, sub-pixel techniques were employed to swiftly extract the position and shape information of the pattern, facilitating rapid pattern recording. We analyzed the rheological characteristics of various ink systems and investigated the impact of different mechanical arm speeds on product printing. We used the shape context matching method to verify that, compared with manual calibration, image calibration improved the similarity of the printed products by 1.81 to 6.29%. Additionally, we demonstrated the effectiveness of image extraction and calibration by successfully printing several different types of patterns.