New algorithm checks bullets at crime scene segment by segment


Credit: CC0 Public Domain

On the morning of March 22, 1915, residents of the small town of West Shelby, New York, woke up in terrible surroundings. A woman, dressed only in a bloodied nightgown, was shot dead in the snow on the threshold of an immigrant farm, Charles Stilov. Across the street, in a farmhouse where Stillau recently worked and where a dead woman was kept, 70-year-old farmer Charles Phelps was found shot dead and unconscious. He died a few hours later.


Upon learning that Stillow lied when he told investigators that he had no weapons, the police arrested him on August 21, 1915. During the trial of Still, the self-proclaimed criminologist Albert Hamilton showed that the nine strokes that he said he found inside the barrel of a 22-gauge Stylo revolver corresponded to nine scratches that he found on bullets of the same caliber at the crime scene. Although Hamilton never showed his testimony to the jury, stating that the results were so technical that only an expert could determine them, Stilow was convicted of murder in the first degree. He was sentenced to electrocution and sent to Sing Sing Prison awaiting execution.

Several people familiar with the case, including the deputy head of the Sing Sing prison, were convinced that Stilou was innocent and that his confession contained words that even a mentally retarded farmer could not understand. The Governor of New York called for a re-investigation just a week before Stillow was to be electrocuted on December 11, 1916. A firearms expert from the New York City Police Department compared the bullets from the scene to the bullets fired from a Still pistol. Even on the eye, the markings on the two sets of bullets did not look the same, but to make sure optician Max Poser studied them under a microscope. Bullets from the scene of the killing could not be fired from Stilow's pistol, he said.

Poser's analysis not only freed Stileva, but also made history one of the first examples of the use of modern forensic methods to detect firearms.

Today medical examiners they still use a microscope developed and improved by two of Poser's colleagues in the 1920s to study bullets or sleeves at a crime scene – metal cylinders that hold powder and bullets until they are fired. Known as a comparison microscope, the device consists of two microscopes connected by an optical bridge.

The split screen of the microscope allows a parallel comparison of tiny scratches or marks on bullets or shells found at a crime scene with markings on bullets or cases fired from a particular weapon. These strips extend to bullets as they are pushed through spiral windings, called rifling, down the barrel of a weapon at high speed and pressure.

The firearms expert adjusts the position of the bullet fired during the test until its banding is better suited to the strip at the crime scene. Thus, the expert can provide his expert opinion on whether the bullets at the crime scene were fired from the same pistol that was produced during the tests.

The method is very successful, but the results of the comparison are subjective and depend on the experience of the expert. Visual comparison does not allow the firearms expert to objectively assess the level of uncertainty in the comparison. For example, what is the probability of obtaining a comparison result if the bullets were actually obtained from the same firearm or from another firearm? Courts now prefer statistical information that, for example, is usually provided by DNA experts when they provide evidence of genetic evidence.

Last year, NIST scientists introduced a computer-based comparison method that can provide this numerical information. The algorithm, known as congruent matching profile segments (CMPS), relies on detailed three-dimensional maps.

“Firearms experts are actually making good comparisons, so it’s not about replacing human judgment with a computer algorithm,” said NIST scientist Robert Thompson, a member of the NIST team. “The algorithm allows you to mathematically evaluate the reliability of the expert’s conclusions.”

It is important to note that instead of comparing the total card or profile of one bullet with another, the algorithm first divides the profile of each bullet at the crime scene into tiny disjoint segments. He then checks to see if any of the individual segments matches any section of the bullet launched during testing.

Segmentation is an important feature, since the bullets at the crime scene are usually deformed or fragmented after ricocheting from a hard surface or rapidly slowing down in the human body. As a result, the grooved groove can be erased, widened or displaced in position. Comparison of the entire profile of such a deformed bullet with the untouched marking of the bullet fired into the water tank can show a low probability of coincidence, even if the bullets could be fired with the same weapon. Searching for suitable objects by segments provides a much more accurate way to compare crime scenes and test markers.

Before the team applied the comparison method, the researchers used image reconstruction methods to “straighten” and display as parallel scratches that were distorted or tilted as bullets deformed. But even after marking on crime scene the bullets are straightened, they may not coincide with the position of the same marking on the test bullets. That's where the CMPS comes in, says PML scientist Johannes Suns. The algorithm takes a small area of ​​the marking on the deformed pool and looks for any place on the test bullets, which may turn out to be a coincidence. The software then estimates how many segments were found in the correct position relative to the markings on the trial pool. The method is based on an older algorithm developed by PML scientist John Song, who compares impressive gun marks on cartridges,

In an initial study that the NIST team reported last December at Forensic Science International, scientists used the CMPS method only to compare undeformed bullets fired from known guns. The team fired 35 9-mm Luger bullets into a water tank of 10 barrels made in series.

On each barrel in the cabinet scratches from bullets are imprinted. Researchers found that CMPS more accurately determines the origin of each bullet than a comparison method that does not divide bullet markings into segments.

In the team’s latest study, published in January by Forensic Science International, researchers first applied the CMPS method to study deformed bullets. The team fired 57 rounds with varying degrees of fragmentation from the same 9 mm pistol into the water tank. Create bullet With fragments of varying degrees of deformation, the researchers directed the gun at a slight angle so that the bullets hit the sides of a large steel tube placed in front of the water tank instead of firing directly into the water.

The team conducted two types of tests using image reconstruction software and the CMPS algorithm. Researchers compared heavily distorted bullet markings to fingerprints on nearly intact reference bullets fired directly into a water tank. They also compared deformed bullets before and after image reconstruction, which leveled distorted markings. Scientists have found that together image reconstruction and CMPS have significantly improved the ability to match markings on deformed bullets with each other and with pristine bullets.

The team is currently planning further research to test the CMPS method. With the freedom — and possibly life — of the accused being in the balance, these studies are crucial to determining whether and when CMPS can be regularly included in the analysis and testimony of firearms experts, says Suns.


How good is the match? Entering statistics into the forensic identification of firearms


Additional Information:
Same Chen et al. Pilot study of the correlation of a deformed bullet, International Forensics (2019). DOI: 10.1016 / j.forsciint.2019.110098

Same Chen et al. Signature correlation using a bullet using the Congruent Matching Profile Segments (CMPS) method International Forensics (2019). DOI: 10.1016 / j.forsciint.2019.109964

This story is republished courtesy of NIST. Read the original story Here,

citation:
The new algorithm checks cartridges at the crime scene by segments (2020, March 27)
restored March 28, 2020
from https://phys.org/news/2020-03-algorithm-crime-scene-bullets-segment.html

This document is protected by copyright. Other than honest deals for private study or research, no
Part may be reproduced without written permission. Content is provided for informational purposes only.





Source link

Leave a Comment