摘要
由于仅依靠原始图像几何中心的坐标和前景区的质心坐标来计算指纹图像的偏移程度,极易丢失指纹的有效信息(如核心点和三角点信息),从而引起后续特征提取与匹配的不准确,为此利用原始图像的几何中心和前景区的质心位置以及提取的奇异点或参考点位置,提出一种指纹图像偏移度计算方法.实验结果表明,该方法能更准确地进行指纹图像的位置偏移度判断.
It would easily miss some useful information, such as the kernel and tri-angle points, if we only relied on the coordinates of the geometric center in the original images and the centroid of foreground area to compute the offset degree of the fingerprint images, which resulted in the subsequent extraction of features and matching being inaccurate. As a result, a calculation method of the offset degree of fingerprint images was proposed by using the geometric center of the original images, the centroid of foreground area and the location of singular points or reference points. The experiment results show that this method can be used for estimating the position offset degree of fingerprint images more accurately.
出处
《空军雷达学院学报》
2009年第3期187-189,193,共4页
Journal of Air Force Radar Academy
关键词
指纹图像
奇异点
偏移度
fingerprint image
singular point
offset degree