Abstract: Image segmentation by thresholding is a usual way in image processing and analysis. With some measures of difference between images, some new methods for image threshold selection are put forward based on the principle that the difference between two parts from an good thresholding segmentation should be big and the differences between original image and two parts are both big. All methods have the simple formula similar with Otsu’s method. Otsu’s method is made based on the principle on many occasions. It verifies that it is correct to make threshold selection method based on that principle. All methods put forward in this paper have been tested to be able to work well for many images.