由于光线分布不均匀或斑块噪音干扰等原因,往往使所要处理的指纹的灰

度值分布缺乏均匀性。在指纹特征自动识别提取过程中,造成许多传统的算法在局部出

现很大的误差。利用方差和均值特征的自动提取方法,首次对不均匀灰度图像进行自适

应分割。然后通过对图像分区域进行不同程度的自适应调整,使具有相同属性的像素单

元具有近似的灰度值分布。调整结果的灰度均匀水平与预先指定的调整精度成正比。这

种调整不仅提高了图像分割的自适应性,而且进一步扩大了一般阈值算法的应用领域。-Gray value distribution of the lack of uniformity due to the light distribution is uneven or plaque noise reasons, tend to make the fingerprint to be processed. Automatic identification of fingerprint feature extraction process, resulting in many traditional algorithm in the local large errors. Automatic extraction method using the variance and the mean characteristics, for the first time on the the uneven grayscale image adaptive segmentation. Then, through adaptive adjustment of the different degrees of the image subregion, so that pixels having the same attribute unit having the approximate distribution of grayscale values. Proportional to the adjustment accuracy of the result of the adjustment of the gray uniform level with a previously specified. This adjustment is not only self-adaptability of image segmentation, but also to further expand the application areas of the general threshold algorithm.