3D Wafer Volume Inspection With Adaptive Image Conformity Control

The patent segments the complex 3D inspection workflow into distinct modular components: image acquisition module, image processing module, and 3D reconstruction module. Each module handles specific tasks independently, allowing the system to maintain high measurement precision while reducing operational complexity through standardized, reusable processing steps.

The patent implements automated parameter optimization that adjusts imaging parameters (such as beam focus, acceleration voltage, and scanning parameters) based on the specific sample characteristics and inspection requirements. This eliminates the need for expert experimentation by automatically finding optimal parameters, thereby maintaining measurement precision while reducing device complexity.

The patent implements self-calibration and self-optimization capabilities where the system automatically adjusts its parameters and processing algorithms based on the sample being inspected. The workflow configuration automatically adapts to different semiconductor structures without requiring expert intervention, enabling routine users to achieve high measurement precision through automated parameter selection and processing optimization.

The patent pre-configures standardized inspection workflows and processing pipelines that are ready to be applied to different sample types. Common inspection scenarios have pre-optimized parameters and processing steps stored in a library, allowing users to select appropriate workflows without performing comprehensive experimentation, thus maintaining precision while improving ease of operation.

The patent implements adaptive sampling strategies that determine the optimal number of cross-section surfaces to mill and image based on the specific inspection requirements and sample characteristics. Instead of always milling a large fixed number of sections, the system uses preliminary analysis to identify regions of interest and adjusts the number of sections accordingly, maintaining high measurement precision while reducing unnecessary milling operations to improve productivity.

The patent implements periodic quality checks and validation steps during the milling and imaging process. Instead of milling all cross-section surfaces continuously, the system periodically verifies measurement accuracy and adjusts the inspection plan accordingly, allowing early termination when sufficient precision is achieved, thus balancing measurement precision with productivity.

The patent implements dynamic raster resolution adjustment that adapts the sampling resolution based on the local features being inspected. In regions with fine critical dimensions, the system automatically reduces the raster size to improve measurement precision, while in regions with larger features, it uses coarser sampling to reduce acquisition time. This dynamic adaptation maintains high precision where needed while minimizing overall inspection time.

The patent applies different sampling resolutions to different regions of the sample based on their importance and feature size. Critical regions with small features receive finer sampling for high precision measurement, while less critical regions use coarser sampling. This localized quality approach optimizes the balance between measurement precision and acquisition time by concentrating computational resources where they are most needed.