Trajectory imaging of low-contrast dynamic objects based on frequency-domain analysis

X-ray imaging allows for the visualization of the internal structure of an object. However, the capture of motion trajectories of weakly absorbing objects in complex backgrounds remains challenging due to low contrast and structural overlap. To address this issue, move contrast X-ray imaging (MCXI) is employed to enhance dynamic features without the use of contrast agents and to provide strong robustness against noise. A fringe-based background model was constructed to simulate and analyze variations in motion trajectories under various influencing factors, including relative contrast, motion speed, background fluctuation frequency, and noise level. Quantitative metrics-such as trajectory intensity, integrity, and signal-to-noise ratio-were used to evaluate the performance of the proposed method. MCXI enables the identification of motion paths and spatial velocity distributions of weakly absorbing objects, even under high-noise conditions, demonstrating high robustness and sensitivity. The effectiveness, sensitivity, and noise resistance of the method were validated through experiments involving oil-water motion. In addition, the integration of the CUDA acceleration algorithm significantly improves the efficiency of processing high-throughput dynamic data.