Optimal Cutting Positioning and Quantitative Slicing Micromanipulation of Wheat Seeds Based on 3D-Phenotypic Feature Estimation
Year: 2026
Type: conference-paper
Abstract: To address the longstanding challenges in wheatseed slicing for breeding-particularly the difficulties of micromanipulation caused by the small size and irregular morphology of wheat seeds, the need for precise localization of the optimal cutting point, reliable clamping, and accurate quantitative cutting, all of which lead to low efficiency and inconsistent slice quality-this study presents an integrated solution framework. First, we introduce a microscopic reconstruction-driven 3D modeling method tailored for small seeds. By embedding a multi-scale feature encoder and an axial attention mechanism into an enhanced PointNet architecture, the approach achieves high-precision point cloud segmentation of wheat seeds, enabling explicit delineation of clamping and cutting regions. This provides a robust basis for accurate extraction of key phenotypic traits from the reconstructed 3D models. Second, a relational model linking cutting volume to 3D phenotypic features is established. Using a contour integral-based inverse cumulative computation, the optimal cutting point is precisely determined, ensuring quantitative consistency in the slicing process. Finally, a stable, sharp-tooth clamping mechanism specifically designed for small and irregular seeds is developed. Critical design parameters are derived, and major structural components are analyzed in detail. Cutting experiments conducted under controlled laboratory conditions further validate the effectiveness and practicality of the proposed mechanism. Test results demonstrate that the reconstructed point cloud achieves 93.71 % coverage of the seed contour, yielding segmentation metrics of 0.91 in precision, 0.83 in recall, and 0.87 in F1 score. Phenotypic feature extraction reaches an average accuracy of 92.71 %. The assembled device exhibits robust clamping performance and high slicing efficiency, achieving a clamping success rate of 91.43 %, a cutting success rate of 95.31 %, and a cutting volume consistently maintained within 6-8 mg, fully meeting the operational requirements for wheat-seed sectioning. Moreover, the germination rate of sliced seeds exceeds 93.44 %. This work provide valuable reference for the design and optimization of clamping mechanisms for small, irregularly shaped seeds.
Authors: Yang Li, Yunfei Chen, Rui Xi, Xiaojun Long
Institution: Shandong Agricultural University
Language: English
Topic: Smart Agriculture and AI
Subfield: Plant Science
Domain: Life Sciences
SDG: Zero hunger
Open Access status: hybrid
OpenAlex ID: W7165766211