Fingerprint enhancement algorithm based on signal analysis and reconstruction

The continuous advancement of biometric technology has positioned research on fingerprint recognition at the forefront, given the significant impact of fingerprint image quality on recognition accuracy. To enhance these images, signal analysis and reconstruction techniques are employed in this paper. Initially, the low-quality fingerprint image is analogized to an interference fringe image, and a corresponding mathematical model (LFMM) is developed. Subsequently, wavelet denoising is applied to the fingerprint image, while a Gaussian low-pass filter is utilized for grayscale compensation, eliminating terms related to grayscale compensation in the model. Next, the Ritz transform serves as an analytical function for signal analysis and reconstruction of the model, generating the fingerprint image’s phase map (LFPM). Finally, Gabor filter enhancement is applied to the original fingerprint image using the fingerprint’s phase map, completing the enhancement process. Experimental results demonstrate the effectiveness of the proposed enhancement algorithm in noise reduction and texture feature recognition enhancement, thus enhancing the accuracy and reliability of the fingerprint identification system. This research carries significant implications for various domains such as security, forensics, border control, and personal identity verification.