Diffraction Image Reconstruction for Flexible Lithography Mask Inspection

The patent creates a computational model that copies and simulates the target illumination conditions rather than physically implementing them. The measured diffraction image data is processed through algorithms that reconstruct what the image would look like under target illumination settings, eliminating the need for complex physical illumination systems while maintaining measurement accuracy.

The patent changes the illumination parameters from physical implementation to computational simulation. By using algorithms to calculate and simulate the effect of different illumination settings on the measured data, the system can evaluate multiple illumination scenarios without physically configuring each one, thereby reducing system complexity while preserving measurement precision.

The patent performs preliminary computational preparation by establishing a computational model that can quickly simulate different illumination conditions. Once the model is prepared, multiple illumination scenarios can be evaluated rapidly through computation rather than physical measurement, enabling versatile inspection across different lithography conditions without proportionally increasing measurement time.

The patent replaces the mechanical process of physically adjusting illumination settings with computational algorithms. Instead of mechanically changing illumination parameters to test different scenarios, the system uses software-based simulation to evaluate multiple illumination conditions, thereby maintaining high versatility while significantly improving inspection productivity.

The patent introduces computational algorithms as intermediaries between the raw diffraction image data and the final inspection results. These algorithms automatically process and interpret the complex diffraction patterns, translating them into meaningful illumination condition evaluations without requiring the user to manually handle the complexity of coherent diffractive imaging data processing.

The system performs self-service processing by automatically analyzing the diffraction image data through built-in algorithms that handle the complexity of coherent diffractive imaging. The system self-calibrates and self-processes the measurements, eliminating the need for user intervention in the complex data processing steps while maintaining high measurement precision.