一种基于多级分块错切的指纹奇异点检测算法

Arithmetic for Singularity Detection Based on Multilevel Block Sizes and Shifting in Fingerprint Images

准确、可靠地检测指纹奇异点(ore点和delta点),对于指纹分类和指纹匹配具有非常重要的意义。针对指纹图像奇异点的精确定位和可靠性判断的难题,提出了一种基于多级分块错切的指纹奇异点检测新方法。首先,对于一枚指纹图像,在同一分块尺寸下进行多次图像错位分块,并且分别在不同的图像错位分块情况下检测指纹的奇异点,得到区域相对集中的奇异点位置的集合,并计算其质心,以精确确定奇异点的位置。然后,在不同的分块尺寸下分别采用平滑和不平滑的方法进行指纹方向场的估计,并分别根据以上方法估计的指纹方向场信息进行指纹奇异点的检测。最后,利用不同情况下检测的指纹奇异点位置相互关联的特性,进行指纹奇异点的精确、可靠检测。该方法利用了多次图像错位分块检测的奇异点位置相对集中和各级分块尺寸下采用不同方法检测的指纹奇异点位置相关联的特性,能够从指纹图像中较精确、可靠地检测出奇异点,对低质量指纹图像具有良好的鲁棒性,在部分典型低质量指纹图像上的实验结果验证了该方法的有效性。

It is very important to detect singularities (core and delta) accurately and reliably for classification and matching of fingerprints. In this paper, we present a new method for singularity detection in fingerprint images to improve accuracy of the position and reliability of the singularity. Firstly, we detect singularities based on block images through shifting position of the whole image time after time at the same block size, get the concentrative region of singularities detected on different positions and compute the centroid of the region to gain the accurate position of singularities. Secondly, we detect singularities based on the orientation field of fingerprint images, which is estimated by different methods (with or without smoothing) and at multi-level block sizes. Finally, on the basis of the corresponding relationship of the singularities detected at multilevel block sizes and by different methods of orientation field estimation, we detect the singularities accurately and reliably. Experiment results show that the method performs well and it is robust to poor quality images.