Machine Learning Based Overlapped Latent Fingerprints Segmentation and Separation

摘要

This paper presents the research in the field of overlapped fingerprints segmentation and separation based on different machine learning techniques. The main contributions of this research are a new method for overlapped latent fingerprints segmentation based on convolutional neural networks, a new method for separation of overlapped fingerprints based on neural networks and a new, large, artificially overlapped fingerprints dataset. Experimental results, obtained using a publicly available datasets and commonly used figures of merit, demonstrate that proposed methods are state-of-art in its field, or competitive to existing methods.