Applied Computer Vision R&D

Group Mission and Programs

The Image Science & Machine Vision (ISMV) Group was established in 1987 to develop technologies that provide human-like decision making capabilities for computers and robots. These methods and systems perform image-based metrology, scene analysis and comprehension, and archival image indexing and management. Today we form one of 18 R&D groups in the Engineering Science & Technology Division at ORNL.

Base Technology and Experience

The group consists of twelve researchers with Electrical Engineering, Computer Science, and Nuclear Engineering backgrounds. We conduct applied computer vision research and development that supports U.S. industry, healthcare, and security through strategic partnerships. Our research encompasses image processing methods that include: image modeling; image transforms; dimensional metrology; object and region segmentation; statistical feature analysis; pattern classification; and content-based indexing and retrieval. We apply these methods to three programmatic areas as described below.

Biomedical Imaging

• Telemedicine and CAD for mammography
• Small Animal high-resolution tomographic medical imaging
• Image Analysis for organ segmentation and phenotype screening

Industrial Inspection

• Semiconductor metrology and inspection
• Manufacturing-based CBIR
• Energy reduction, waste mitigation, and QC

National Security

• Threat object detection in x-ray imagery
• Electronic fingerprint QC and uniqueness
• Nuclear materials verification / identification
• Geographic information analysis

Research Facilities

• Digital Holography Microscopy Laboratory
• Small Animal Imaging Laboratory
• Oak Ridge Web Test Bed Laboratory
• Video Analysis Laboratory