【摘要】 分析了体数据小波变换的空间 -频域特征 ,提出了一种构造体数据层次模型的新方法。新方法将小波变换的结果重组为有效位置体 ,并采用基于系数尺度立方体的细化规则 ,通过自上而下逐尺度加入高频细节的策略来构造体数据的层次模型。与现有方法不同 ,新方法直接根据小波系数来判定体数据局部区域的复杂度 ,避免了费时的小波逆变换 ,简化了模型的构造 ,节省了时间。实验证明 ,新方法性能良好 ,简单易行 ,适合对大容量规则数据集进行交互考察与分析更多还原

【Abstract】 A new method for constructing hierarchical models of vo lume data is proposed, based on examining precisely the spatio-frequency locali zation characte-ristics of 3D wavelet transform. The wavelet coefficients are f irst rearrange into the effective position volume, then,the local complexi ties of the volume data are determined by directly using wavelet coefficients an d th e volume data approximation with the coefficient-scale-cube-based refinement rule. So the time-consuming inverse wavelet transformation is avoided and the construction process simplified. The simulational results of the computer for a huma n being data show that the algorithm is simple and effective. It is suitable to visualize the large regular volume data interactively.更多还原