因为传统分割算法对图像细节特征信息不敏感，分割效果不理想，提出了一种确定阈值的改进方法。该算法在阈值图像法的基础上，采用迭代计算平均值的方法，动态地获得每个像素点位置的最佳阈值，以获取一幅与原图像大小相同的阈值图像，利用该阈值图像对原静脉图像进行二值化处理，增强了图像的局部细节特征信息。仿真实验证明了该改进算法的有效性。

Traditional algorithms of hand vein image segmentation are insensitive to image detail and have poor segmentation effect. A new improvement method based on threshold image was put forward. The key of this new method is iterative calculations instead of calculating the mean of every plate. It divides the image into several parts, and each part uses an optimal threshold scheme to get a threshold image. Its size is the same with the original vein image. Then the partial information can be enhanced through binary processing of the image. The simulation experiment showed that the proposed method is effective.