In this correspondence, we present a general recursive approach for im
age segmentation by extending Otsu's method. The new approach has been
implemented in the scope of document images, specifically real-life b
ank checks. This approach segments the brightest homogeneous object fr
om a given image at each recursion, leaving only the darkest homogeneo
us object after the last recursion. The major steps of the new techniq
ue and the experimental results that illustrate the importance and the
usefulness of the new approach for the specified class of document im
ages of bank checks will he presented.