Abstract
To find the need of images from Internet is a heavy work, image annotation techniques can be used to help us quickly find the required images, the traditional image annotation method can't between the image and text classification, cannot achieve most web users demand, this paper proposes an image semantic annotation technology automatically, based on the SVM technology. The first we map image visual feature to one or more images of rough concept by using annotation, then we preprocess web page text information, lastly, we selected high similarity of keywords as image annotation. From the experimental effect it can effectively improve the image tagging precision.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call