We proposed a hybrid CSG and B-rep method to reconstruct 3D CAD models from multi-view engineering drawings. Visual cognition is introduced into understanding engineering drawings. A visual reasoning engine is implemented to deal with the ambiguous solutions in its geometric an topologic interpretation. We also investigate how to interpret section drawings in 3D reconstruction from projections. Based on the "divide and conquer" strategy, we further provide a systematic explanation of why this approach works, and implement an experiment system to demonstrate its effectiveness and efficiency. Moreover, we explored the relevant visual cognition theory, computational models and algorithm frameworks regarding combining visual cognition with the understanding of engineering drawings. The major achievement of this project was published in a well-known international journal in CAD filed, "Computer-aided Design". The relevant paper is cited by SCI. The research output of this project has potential applications in facilitating the product modeling in CAD, the reuse of design, and human computer interaction.
本项目研究工程视图理解的认知理论和可计算模型,探索和研究与人的理解较为一致的工程视图理解的新方法和新途径,并通过原型系统来验证和实现上述模型和方法。本项目的研究对计算机辅助设计、计算机集成制造和认知科学等学科领域的发展起重要推动作用,使我国在该领域的研究和应用跻身世界先进水平。
{{i.achievement_title}}
数据更新时间:2023-05-31
基于SSVEP 直接脑控机器人方向和速度研究
伴有轻度认知障碍的帕金森病~(18)F-FDG PET的统计参数图分析
物联网中区块链技术的应用与挑战
人工智能技术在矿工不安全行为识别中的融合应用
连续视程人工晶状体植入术后残余散光对视觉质量的影响
基于NAM的动态视觉信息认知理解方法研究
基于视觉认知理论的图像层次化语义理解研究
基于认知计算和信息挖掘的多波段夜视图像融合技术
融合先验建模和深度学习的自然场景视觉理解研究