Abstract: In this paper, we propose an incomplete image annotation algorithm based on spatial
information. We first select the minimum rectangular region of the incomplete part of the picture, and
then segment all the images by extending the selected region as the segmental directrix. We calculate
the similarity of pictures by using the image segmentation sub-block as the fundamental unit. We and
use the spatial structure information of image segmentation sub-block to complete the automatic
image annotation. The experimental results proved that our method could reduce the negative impact
of image content understanding which is influenced by the incomplete part of the picture. Our method
could also improve the recall ratio and precision ratio of image annotation. It could better reflect the
information contained in the whole image.

Key words: spatial information, incomplete image, image annotation