Quantitative/Statistical Approach to Bullet-to-Firearm Identification with Consecutively Manufactured Barrels
Efforts to use objective image comparison and bullet scanning technologies to distinguish bullets from consecutively manufactured handgun barrels from two manufacturers gave mixed results. The ability of a technology to reliably distinguish between matching and non-matching bullets, where the non-matching bullets were as close in pattern to the matching ones as is probably possible, would provide evidence that the distinctions could be made ''objectively'', and independently of human eyes. That evidence is identical or very close to what seems to be needed to satisfy Daubert standards. It is fair to say that the FTI IBIS image comparison technology correctly distinguished between all the Springfield barrel bullets, and between most but not all of the HiPoint barrel bullets. In the HiPoint cases that were not distinguished 100% of the time, they would he distinguished correctly at least 83% of the time. These results, although obviously limited to the materials used in the comparisons, provide strong evidence that barrel-to-bullet matching is objectively reliable. The results with SciClops were less compelling. The results do not mean that bullet-to-barrel matching is not objectively reliable--rather, they mean that this version of the particular technology could not quite distinguish between these extremely similar yet different bullets as well as the image comparison technology did. In a number of cases, the numerical results made the correct distinctions, although they were close to one another. It is hard to say from this data that this technology differs in its ability to make distinctions between the manufacturers, because the results are very similar with both. The human examiner results were as expected. We did not expect any misidentifications, and there were not any. It would have been preferable to have a higher return rate, and thus more comparisons in the overall sample. As noted, the ''consecutively manufactured barrel exercise'' has been done before, with the same outcome.
- Research Organization:
- Ames Lab., Ames, IA (United States)
- Sponsoring Organization:
- USDOE
- DOE Contract Number:
- W-7405-Eng-82
- OSTI ID:
- 892804
- Report Number(s):
- IS-5170; TRN: US200623%%535
- Country of Publication:
- United States
- Language:
- English
Similar Records
Adrenal imaging with iodomethyl-norcholesterol (I-131) in primary aldosteronism
Low Count and Background Radionuclides Analysis - 20488