Abstract: Traditional 2D Otsus method supposes that the sum of probabilities of main- diagonal distinct in 2D histogram is approximately one. This papers experiments and theory prove that the sum of probabilities of counter-diagonal distinct in 2D histogram is not always very small and this could not be neglected. Therefore, the assumption mentioned above in 2D Otsus method is inadequately reasonable. In this paper, an improved 2D Otsus segmentation method and recursive algorithm are proposed. By separately calculates probabilities of objects area and background area in 2D histogram main-diagonal distinct, modified method is acquired. Experimental results show that proposed method can obtain better segmentation effect than the traditional 2D Otsus method. The computation complexity of improved 2D Otsus method corresponds to traditional 2D Otsus method.