Structure-Aware Spectral Reconstruction via Observation Window Compensation

This work presents a structure-aware reconstruction framework for recovering suppressed high-order components under observation window distortion.

In many imaging and signal acquisition systems, the observation process acts as a non-uniform filter that attenuates specific structural components, leading to incomplete or biased reconstruction. We model the observation process as a weighted projection over structural components and introduce a window compensation approach to invert such distortions.

By estimating the observation window and applying component-wise correction, the proposed method enables recovery of hidden structural information that is otherwise inaccessible from single observations.

Numerical simulations demonstrate significant improvements in reconstruction accuracy, robustness under noise, and tolerance to window estimation errors. The framework provides a general method for structure-level reconstruction in multi-component systems and can be extended to imaging, signal processing, and related applications.