Abstract: Cephalometry in orthodontics is gradually developing from 2D to 3D. As a crucial basis for orthodontics, cephalometric radiographs are limited to 2D and inevitably have problems such as ambiguity, distortion, and ghosting, while measurement and analysis methods based on 3D images, such as CBCT, are not perfect. The existing research on 3D medical imaging primarily focuses on basic tasks such as denoising and reconstruction, and less research on 2D image replacement. For this reason, this paper proposes a CBCT reorientation method based on 3D landmarks and a 2D cephalometric simulation visualization method. First, specific landmarks are predicted on 3D images, and the head coordinate system is established according to the corresponding relationship between landmarks and head anatomical structure to redirect CBCT images and align to form unified standardized CBCT data; Then, the imaging process of 2D cephalometric radiographs is simulated, and the post-processing methods such as nonlinear mapping and region enhancement are applied. Based on the redirected CBCT, the simulated 2D cephalometric radiographs convenient for observation and measurement are generated, which can be directly used in the existing 2D cephalometric automatic and manual methods. This paper verifies the proposed method's advancement and effectiveness through many experiments and user studies. The results can replace 2D images for cephalometric analysis, and provide a visual basis for orthodontics.