Optimization and Statistical Testing of Experimental and Computer Methodologies for Evaluation of the LIT-MIA Thermal Imaging Method for Recovery of Defaced Serial Numbers

As submitted by the proposer:

Previous work by the researchers demonstrated that Lock-in Infrared Thermal Imaging, coupled with the multivariate image analysis processing methods of moving average filtering (MAF), principal component analysis (PCA), Zernicke moment analysis (ZMA), and a fusion of similarity merit measurements showed promise as a versatile, non-destructive method for the recovery of defaced serial numbers. The researchers called the technique the LIT-MIA method.

This methodology was employed to successfully recover defaced serial and VIN numbers that were stamped or laser engraved into a wide variety of materials of various shapes and sizes. The multivariate processing of the thermal images provides the information necessary for a comparison of the recovered images with a library of numbers through similarity merit measurements, and shows promise for minimizing cognitive bias regarding the "best-matched" identity of the defaced numbers.

Perhaps the greatest value of the LIT-MIT thermal imaging technique is its non-destructive versatility as a quickly-implemented investigative tool to either help return stolen property to its owner or to tentatively implicate someone in a crime. Since it is non-destructive, the evidence can later be more thoroughly analyzed in a crime lab setting in conjunction with more traditional methods or perhaps as a standalone method. Thus, it may be used for cases that never make it to court or to bolster evidence that is being presented in court.

The LIT-MIT method requires additional study and validation before the method can be designated for routine or specialized serial number recovery including more "field testing" on defaced samples that have been subjected to a "controlled defacement" as well as with real-world samples. A black box approach will be initiated where defaced samples are obtained from an independent source.

Defaced samples will first be recovered in our lab using the LIT-MIA methods and subsequently by another lab that uses a more conventional serial number recovery method. The nondestructive nature of the thermal imaging method makes this sequential analysis possible.

Note: This project contains a research and/or development component, as defined in applicable law, and complies with Part 200 Uniform Requirements - 2 CFR 200.210(a)(14).

ca/ncf