Integrated Imaging System Enhances Detection Capabilities

Friday, August 01 2008

A technology that picks out minute abnormalities in forged documents or suspicious biological spores has been developed by Chem - Image Corp. of Pittsburgh, PA through funding from the Missile Defense Agency (MDA).

The technology integrates two specialized types of imaging — luminescence and Raman — into spectroscopes to provide high-throughput screening for defects or anomalies in microscopic semiconductor materials such as gallium nitride (GaN). Abnormalities in large, visible items such as paper documents also can be screened by these spectroscopes as part of forensic trace evidence testing. MDA predecessor BMDO originally funded the technology for GaN screening through an SBIR (Small Business Innovation Research) Phase I contract in 2000. Though ChemImage received only Phase I funding from MDA, company scientists were determined to press on and develop a commercially available system. ChemImage soon integrated the fundamental concepts addressed in the Phase I contract into three systems that are now commercially available: the Falcon™, Condor™, and CI Vision™ multimodal hyperspectral imaging systems.

How it Works

Scientists use luminescence imaging — or ultraviolet excitation — to find or target the anomaly or defect of a particular semiconductor or material. Raman imaging then provides specific identification of these undesirable occurrences. ChemImage has designed special software that integrates luminescence and Raman into a dynamic imaging system to find and identify an abnormality.

Detection of a defect or anomaly (such as dislocations, micropipes, and low-angle boundaries) in a GaN semiconductor or other object starts with a luminescence imaging device that floods ultraviolet (UV) light onto the material with wavelengths between 250 and 350 nanometers (nm). The UV illumination causes the device or object to glow, revealing any abnormalities. An imaging spectrometer then captures images of the material being examined, with images gathered at wavelengths of greater than 350 nm. After the images are taken, ChemImage applies a technique known as multivariate image analysis. This technique processes every pixel in each of the images independently (see photos).

Each pixel has a luminescent spectrum associated with it. The imaging system generates a digital image showing all aspects of the object in question — including defect density — once the individual pixels are processed. After the luminescence system completes its digital imaging of a particular object, the data are transferred within the platform to the Raman device, which confirms defect types or other anomalies through what is known as hyperspectral imaging. Hyper - spectral imaging provides quantitative analysis of the chemical makeup of an abnormality in a material.