Dactylogram Classification With CNN Confidence Verification

The patent replaces traditional manual or rule-based classification methods with a convolutional neural network (CNN) based automated classification system. The CNN processes fingerprint images through multiple convolutional layers that automatically learn and extract features, substituting the mechanical/manual classification process with an intelligent automated system that achieves higher accuracy for diverse fingerprint types including digital, partial, and full palm prints.

The patent introduces a verification module as an intermediary between the CNN classification output and the final classification result. This verification module checks the confidence scores and validates the classification, acting as a mediator that ensures accuracy while maintaining system reliability. The intermediary layer helps resolve the contradiction by adding a layer of validation without completely redesigning the entire system.

The patent implements a self-service classification system where the CNN automatically classifies fingerprint images without requiring manual intervention from operators. The system performs feature extraction, classification, and verification autonomously, eliminating operator errors while maintaining ease of operation through automated processing. The self-service approach resolves the contradiction by making the system both reliable and operationally simple.

The patent incorporates a feedback mechanism through the verification module that monitors classification confidence scores and can request re-classification or validation when uncertainty is detected. This feedback loop continuously improves reliability by learning from verification results while maintaining operational simplicity through automated decision-making, reducing the need for manual oversight.

The patent applies partial verification action by focusing the verification module on checking only the most critical aspects of classification confidence and consistency rather than performing exhaustive verification of all possible fingerprint characteristics. This selective verification approach ensures database integrity for the most important cases while reducing overall processing time by avoiding unnecessary comprehensive checks on every fingerprint.

The patent performs preliminary classification using the CNN before applying verification, allowing the system to quickly process and classify the majority of fingerprints that are clearly identifiable. The verification is then applied selectively to cases where the preliminary classification shows uncertainty or lower confidence scores, thus maintaining database integrity while minimizing the time spent on verification for clearly classified fingerprints.