The concept of the two-dimensional bound histogram is proposed,and an automatic threshold selection method based on that histogram is presented.In the method,the bound set of an image and its corresponding two-dimensional bound histogram are constructed and the gray-level threshold for segmentation is determined by the Otsu algorithm based on that histogram.In order to prove the validity of the method,it was applied to thresholding of the image of a small underwater target.The Otsu algorithm based on the one-dimensional bound histogram was also used.The results show that the former,the proposed method,is well applicable to the image with a non-ideal bimodal histogram,and has better antinoise performance nad segmental effect than the latter.