【摘要】 指纹的方向信息是指纹图像处理中非常重要的信息,指纹方向场的获取以及校正直接影响指纹自动识别的性能。本文提出了一种基于指纹方向场数学模型的方向校正方法,根据指纹中的奇异点性质和分布,建立起与对应指纹方向场的最优方向场模型;然后,根据模型与实际指纹方向场的残差对指纹方向进行校正。由于算法利用指纹整体的拓扑信息对指纹噪声方向进行校正,具有很强的抗干扰能力,较适合劣质指纹图像处理。更多还原

【Abstract】 Orientation information is very important in the fingerprint image processing. The estimation and correction of the orientation field affect the automatic fingerprint recognition directly. This paper proposed an algorithm of orientation correction based on the fingerprint mathematic model. First an optimal fingerprint model was established via the singular points and the coarse orientation field of the fingerprint image. Then an orientation correction was applied on the residual error of the model. As the global topological information is applied, the algorithm performs well in the noise resistant especially in the poor quality fingerprint images.更多还原