【摘要】 针对目前自适应分层算法中仍存在着算法复杂度高、无法减弱阶梯效应和识别模型的一些细节特征等问题，文中对基于三角面片法向量的自适应分层算法进行了改进。该算法为消除阶梯效应，以三角面片与切片平面间的夹角为判断依据；为准确识别模型特征线和面，以二面角和特征线阈值为判断依据；为更准确地识别模型特征点，文中引入顶角加权改进了Taubin顶点法向量的估算公式，将其作为新的判断依据。最终综合考虑这3个判断依据，提出一种有效保留STL模型特征的改进自适应分层算法。通过打印试验表明：相较于传统的基于三角面片法向量和原Taubin顶点法向量的自适应分层算法，文中改进分层算法下的模型成形精度分别提升了39%和13%,能够更好地兼容成形精度和效率。更多还原

【Abstract】 Since the adaptive layering algorithms suffer high complexity and inability to weaken the step effect and identify some detailed features of the model, in this article efforts are made to improve the adaptive layering algorithm based on the triangular facet normal vector. With the help of the algorithm, the angle between the triangular facet and the slice plane is used as the judgment criterion to eliminate the step effect. The dihedral angle and the feature line threshold are used as the judgment criteria to accurately identify the model’s feature lines and surfaces. The vertex angle weighting is introduced to improve the estimation formula of the Taubin vertex normal vector, which is used as a new judgment criterion to more accurately identify the model’s feature points. Finally, with these three criteria taken into consideration, an improved adaptive layering algorithm with the STL model features effectively preserved is proposed. A series of printing experiment show that compared with the traditional adaptive layering algorithms based on the triangular facet normal vector and the original Taubin vertex normal vector, this improved layering algorithm enhances the model’s forming accuracy by 39% and 13%, respectively, with better compatibility of forming accuracy and efficiency.更多还原