20240054634. RESIDUE CLASSIFICATION FROM MACHINE LEARNING BASED PROCESSING OF SUBSTRATE IMAGES simplified abstract (Applied Materials, Inc)
RESIDUE CLASSIFICATION FROM MACHINE LEARNING BASED PROCESSING OF SUBSTRATE IMAGES
Organization Name
Inventor(s)
Sivakumar Dhandapani of San Jose CA (US)
Arash Alahgholipouromrani of San Jose CA (US)
Dominic J. Benvegnu of La Honda CA (US)
Kiran Lall Shrestha of San Jose CA (US)
RESIDUE CLASSIFICATION FROM MACHINE LEARNING BASED PROCESSING OF SUBSTRATE IMAGES - A simplified explanation of the abstract
This abstract first appeared for US patent application 20240054634 titled 'RESIDUE CLASSIFICATION FROM MACHINE LEARNING BASED PROCESSING OF SUBSTRATE IMAGES
Simplified Explanation
The abstract describes a neural network trained for substrate residue classification by obtaining ground truth residue measurements of a calibration substrate and converting color images of die regions to residue level measurements.
- Ground truth residue measurements obtained for a calibration substrate
- Color images of die regions converted to residue level measurements by neural network
- Training process for substrate residue classification system
Potential Applications
- Quality control in semiconductor manufacturing
- Automated inspection of substrates for residue levels
Problems Solved
- Manual inspection and measurement of substrate residues
- Inconsistencies in residue classification across different substrates
Benefits
- Increased accuracy and efficiency in residue classification
- Reduction in human error in measurement process
- Cost savings through automation of inspection process
Original Abstract Submitted
a neural network is trained for use in a substrate residue classification system by obtaining ground truth residue level measurements of a top layer of a calibration substrate at a plurality of locations, each location at a defined position for a die being fabricated on the substrate. a plurality of color images of the calibration substrate are obtained, each color image corresponding to a region for a die being fabricated on the substrate. a neural network is trained to convert color images of die regions from an in-line substrate imager to residue level measurements for the top layer in the die region.
