In this paper, a shape saliency measure for only shape feature of each object in the image is described. Instead
biologically-inspired bottom-up Itti model, the dissimilarity is measured by the shape feature. And, Fourier descriptor is
used for measuring dissimilarity in this paper. In the model, the object is determined as a salient region, when it is far
different from others. Different value of the saliency is ranged to generate a saliency map. It is shown that the attention
shift processing can be recorded. Some results from psychological images and remote sensing images are shown and
discussed in the paper.