本文回顾了图像分离中一些典型阈值的选择技术,提出了一种自适应阈值选择方法。该方法是基于对所给图像景物和背景信息量的测量而定,其过程很简单。作为这种方法的进一步发展,对可变或阈值选择方法进行了研究,它是将一幅完整图像分割成若干子图像进行综合处理。最后,几个实验结果证实了本方法的有效性. This paper reviews some typical threshold selection techniques in image separation and proposes an adaptive threshold selection method. The method is based on measuring the amount of scene and background information given to the image, and the process is simple. As a further development of this method, a variable or threshold selection method is studied, which is to divide a complete image into several sub-images for comprehensive processing. Finally, several experimental results confirm the effectiveness of this method.