I've been working on a basic pattern recognition and verification course which started as a project for a non criminal automated system. While I was preparing the course I realized that if I did it on line I could reach a broader audience. The course could be used by law enforcement as a basic foundation course to develop latent examiners, individuals working on automated fingerprint systems or even law offices so they could actually evaluate the fingerprint and know when they need a fingerprint expert.
In today's world dominated by the automated fingerprint identification systems training has been simplified, we no longer need to know the full fingerprint classification formula. While the classification formula has been simplified the ability to determine pattern should not be simplified. As all experienced latent or tenprint examiners know; the ability to distinguish pattern type to include the sub-groupings is critical in learning and understanding fingerprint basics. Many automated systems use pattern as a filter or binning tool to increase the speed and accuracy of the system. The fingerprint examiner has to understand the difference between a left slant loop and a central pocket loop whorl or a left slant loop and a spoiled loop tented arch. If the differences aren't understood the database gets corrupted and system accuracy will be diminished. People tend to concentrate on the experienced latent fingerprint examiner and fail to realize that the database the examiner uses is updated and maintained by individuals with far less experience and training.
A good basic pattern recognition/verification course is essential in developing a good database as well as latent examiners, preventing fingerprint errors and creating reliable search databases.
Bob McAuley
Dir. Operations/Training
Forensic Biometric Identification Solutions LLC.
Showing posts with label biometric automated fingerprint identification systems expert fingerprint court testimony fingerprint latent Latent fingerprint tenprint fingerprint. Show all posts
Showing posts with label biometric automated fingerprint identification systems expert fingerprint court testimony fingerprint latent Latent fingerprint tenprint fingerprint. Show all posts
Saturday, August 22, 2009
Wednesday, August 5, 2009
How to obtain more reliable fingerprint evidence
Read the 2007 paper by Roger Koppl "CSI for Real: How to Improve Forensics Science". The paper is well written, thoughtful and accurate. The paper presents some very good ideas on how to improve the system. What caught my eye was the study of proficiency tests conducted by the Forensic Sciences Foundation (FSF) and Collaborative Testing Services (CTS) from 1978-1991.
In the fingerprint test they “attempted to simulate actual conditions by creating smudged, elongated, compressed and other irregular latent print specimens.” The best results were for
the group that included finger and palm prints, the rate of false identifications for the period was 2 percent. The researchers estimate that fingerprints appear in roughly 7 percent of felony case filings each year. Roughly 924,700 adults were convicted of a felony in State courts in 2000” and another 77,000 in U.S. district courts. These numbers suggest that about a million felony cases are brought to trial and concluded each year. Approximately 70,000 involve fingerprint evidence. It can reasonably be inferred, that of these 70,000 cases, at least 2 percent or 1,400, involve a false identification. The identification of a latent print lifted from a crime scene with an inked print from a suspect will usually produce a conviction.
The 2% figure is consider low by some, other studies have produced greater percentages of error, regardless, even a small percentage of errors produce a significant number of incorrect identifications. While not discussed in the article the number of missed identifications is significantly larger and is another cause of individuals being wrongly convicted.
While Prof. Koppl has some excellent suggestions, in the interim it is the responsibility of defense counsel to insure fingerprint evidence is accurate and reliable. The adversarial system of justice allows even fingerprint evidence to be questioned and tested to insure accuracy. In today's electronic world fingerprint evidence being brought to trial in NY, can be evaluated by an independent examiner in NE or any other state. The costs, because travel and shipping are no longer an issue, are reasonable enough that every defendant should be able to have an evaluation of the fingerprint evidence.
Bob McAuley
Dir. Operations/Training
Forensic Biometric Identification Solutions LLC.
In the fingerprint test they “attempted to simulate actual conditions by creating smudged, elongated, compressed and other irregular latent print specimens.” The best results were for
the group that included finger and palm prints, the rate of false identifications for the period was 2 percent. The researchers estimate that fingerprints appear in roughly 7 percent of felony case filings each year. Roughly 924,700 adults were convicted of a felony in State courts in 2000” and another 77,000 in U.S. district courts. These numbers suggest that about a million felony cases are brought to trial and concluded each year. Approximately 70,000 involve fingerprint evidence. It can reasonably be inferred, that of these 70,000 cases, at least 2 percent or 1,400, involve a false identification. The identification of a latent print lifted from a crime scene with an inked print from a suspect will usually produce a conviction.
The 2% figure is consider low by some, other studies have produced greater percentages of error, regardless, even a small percentage of errors produce a significant number of incorrect identifications. While not discussed in the article the number of missed identifications is significantly larger and is another cause of individuals being wrongly convicted.
While Prof. Koppl has some excellent suggestions, in the interim it is the responsibility of defense counsel to insure fingerprint evidence is accurate and reliable. The adversarial system of justice allows even fingerprint evidence to be questioned and tested to insure accuracy. In today's electronic world fingerprint evidence being brought to trial in NY, can be evaluated by an independent examiner in NE or any other state. The costs, because travel and shipping are no longer an issue, are reasonable enough that every defendant should be able to have an evaluation of the fingerprint evidence.
Bob McAuley
Dir. Operations/Training
Forensic Biometric Identification Solutions LLC.
Subscribe to:
Posts (Atom)
Posts
