Fingerprint Authentication Device Singular Point Extraction

The patent applies preliminary action by using machine learning to pre-process and identify the interest region containing the singular point before traditional authentication processing. The singular point determiner circuit uses trained models to detect and locate the singular point (core or delta) in advance, allowing the system to extract only the relevant interest region for authentication. This preliminary identification of the interest region based on singular point location enables faster processing while improving reliability by focusing on the most informative areas of the fingerprint image, avoiding noise interference from other regions.

The patent applies segmentation by dividing the fingerprint image processing into distinct functional modules: a singular point determiner circuit that uses machine learning to detect the singular point, an interest region extractor that isolates the relevant area based on the singular point location, and an authenticator that performs the actual authentication. This segmentation allows the complex machine learning task to be confined to a dedicated module, while other parts of the system remain relatively simple. The singular point determiner can be implemented as a separate processing unit or integrated into the existing image processor, enabling parallel processing and improving authentication speed without overwhelming the entire device with complexity.