Isaret i Takibi ile PnP Tabanli 6DoF Poz Tahmini ve CFD Sim lasyon Kar sila stirmasi

This study employs the Perspective-n-Point (PnP) method for estimating the six degrees of freedom (6DoF) pose of a store. The 6DoF framework encompasses both translational and rotational components along three axes. A Computational Fluid Dynamics (CFD)-based simulation provides the reference 6DoF motion data. Camera calibration is performed using images of a chessboard pattern captured from various angles in order to determine the intrinsic parameters. The store is equipped with markers, whose two-dimensional positions are tracked across video frames. The PnP algorithm is applied using the tracked markers and the calibrated camera parameters to estimate the pose of the store, and the resulting estimates are refined using an Exponential Moving Average (EMA) filter. The accuracy of the pose estimation is evaluated by comparing the results against CFD data using the Root Mean Square Error (RMSE) metric. This work systematically analyzes the performance of the PnP-based pose estimation approach by benchmarking it against high-fidelity simulation data.