The article examines a promising statistical model that can be used to optimize the workflow of fingerprints verification by forecasting the need to find some unknown trace in the Automated Fingerprint Identification System - AFIS. The proposed model rules out the need for careful study of poor quality fingerprints that involves additional consumption of resources. It is argued that this model can be used to effectively control the workflow and the workload by classifying the above-mentioned traces depending on the quality and quantity of information they contain. This allows experts to choose the optimum processes for verification of each fingerprint.
|
