| Imagu is a developer of cutting-edge image analysis software for process yield control and enhancement. Deployed by high-end semiconductor vendors, our software elevate the accuracy of Automatic Defect Review (ADR) and Automatic Defect Classification (ADC) while reducing development risks and costs.
- Improved accuracy of ADR
- Improved accuracy and purity of ADC
- Improved detection of user defined Defects Of Interest (DOI).
- Reduction of nuisance defects allows for higher sensitivity and better defect detection.
Imagu software enables real ADC, which is classification of defects to groups of interest, and not only nuisance filtering. Standard ADC and ADR software suffer from four major hurdles:
1. Inspection software typically detect changes within and between images.
2. ADC is extremely challenging.
- Reducing nuisance defects often means reducing sensitivity
- There are many changes on the wafers that can be tolerated, and are not DOI
- There are “events” on the wafer that, depending on their location or level, can be either defects or nuisance
3. ADR is complicated and often not sufficiently reliable.
- Typically used for nuisance reduction and not for defect classification
- Long setting time
- Lot variations require modifications
- Frequently, purity and accuracy rates are not satisfying
4. The development process is extremely costly: Development is lengthy, often taking years to complete, and the chance of failure is considerable.
- Requires a large search window or FOV which contradicts the ability to recognize the defect and to classify it
Imagu’s solution effectively overcomes these hurdles, enabling full automation through higher standardization and increased system optimization, raising semiconductor ADR and ADC to a new level. For more information on our technology please refer to the Technology section of our web site.
- Precise image segmentation
- Highly sensitive defect detection, without detecting nuisance defects
- Aggregation of data from multiple data sources
- Classification of different elements such as layers, defects and complex patterns
- Classification of defects in multi layered images
- Complements existing comparison / change-based detection software by increasing process accuracy & yield purity
- Single image (i.e. non reference based) detection
- Support for all image sources including Optical Microscopy and electronic Microscopy
- Substantial reduction of development risks and costs and time to market: feasibility or “proof of concept” is reached within weeks; development of new applications, in accordance with customer requirements and goals, can be achieved within just a few months