Hierarchical Bricking for Semiconductor Image Data Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Processing large 3-D image data sets of semiconductor components is complex and memory-intensive due to their high resolution and size, making it difficult to efficiently identify defects and reduce process costs during wafer fabrication.
Innovation Solution
Applying a hierarchical bricking technique to decompose image data sets into bricks of varying sizes across multiple hierarchical levels, allowing for efficient processing and navigation through the data set while maintaining context information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional processing techniques are used on large 3-D image data sets, then measurement precision can be maintained, but processing time and memory requirements become excessively high
Solution Approach 1:
The patent applies hierarchical bricking to divide the large 3-D image data set into multiple bricks of varying sizes across different hierarchical levels. This segmentation allows the processing system to work with smaller, manageable subsets of data rather than the entire data set at once, significantly reducing processing time while maintaining the ability to detect defects at appropriate resolution levels.
Solution Approach 2:
The patent introduces a hierarchical level dimension to the traditional processing approach. By organizing bricks across multiple hierarchical levels (from coarse to fine resolution), the system adds a temporal/organizational dimension to processing, enabling efficient navigation through the data set without requiring all data to be processed simultaneously at maximum resolution.
2Reliability
If conventional processing techniques are used on large 3-D image data sets, then complete defect analysis can be performed, but memory requirements become excessively high
Solution Approach 1:
By segmenting the image data set into a hierarchical brick structure, the patent enables the system to load and process only the necessary bricks into memory at any given time. This reduces peak memory requirements while maintaining defect detection reliability, as the hierarchical organization allows selective access to relevant data regions without requiring the entire data set to be resident in memory simultaneously.
Solution Approach 2:
The hierarchical bricking creates a nested structure where finer-resolution bricks are contained within coarser-resolution bricks across multiple hierarchical levels. This nesting allows the system to navigate from coarse to fine levels, loading only the necessary detail into memory based on the region of interest, thereby reducing overall memory requirements while preserving complete defect analysis capability.
3Measurement precision
If high-resolution processing is applied to entire large image data sets, then defect detection precision is improved, but processing complexity increases
Solution Approach 1:
The patent applies local quality by processing different regions of the image data set at different hierarchical levels based on their importance and defect likelihood. Rather than uniformly processing the entire data set at maximum resolution, the system can focus high-resolution processing only on regions of interest identified at coarser levels, reducing overall processing complexity while maintaining precision where it matters most.
Solution Approach 2:
The hierarchical bricking structure enables preliminary action by allowing the system to first process coarser-resolution bricks to identify potential regions of interest or anomalies. Only after this preliminary screening does the system apply high-resolution processing to specific bricks containing suspected defects, thereby reducing overall processing complexity while maintaining defect detection precision.
Data Source
AI summary
A method includes obtaining an image data set that depicts semiconductor components, and applying a hierarchical bricking to the image data set. In this case, the bricking includes a plurality of bricks on a plurality of hierarchical levels. The bricks on different hierarchical levels have different image element sizes of corresponding image elements.