Segmented image regions may still have multi-objects, so the extractive visual features may not effectively describe the corresponding regions. In order to overcome the above problem, an image annotation model based on multi-scale salient region is proposed. Our model can improve the object descriptions of images and image regions, and then the overall performance of image annotation could be improved. In future work, we will consider how to improve the effects of object segmentation in automatic image annotation by combining the saliency maps with local features, and then improve the overall performance of automatic image annotation.
展开▼