Abstract: The Fisher information (FI) measure is an important concept in statistical estimation theory and information theory. However, it has received relatively little consideration in image processing. In this paper, a novel algorithm is developed based on the nonparametric FI measure. The proposed method determines the optimal threshold based on the FI measure by maximizing the measure of the separability of the resultant classes over all of the gray levels. This method is compared with several classic thresholding methods on a variety of images, including some nondestructive testing (NDT) images and text document images. The experimental results show the effectiveness of the new method.