Texture-based semantic annotation of insect images
The traditional artificial annotation of insect images is time-consuming and laborious. Moreover, for subjective factors, the results are always not uniform, which results in unsatisfactory results of the subsequent retrieval of insect images. So in this paper, a new method of image annotation was proposed, based on texture feature of insect images. First it introduced the feature extraction of the objects from insects images; And then elaborated on the implementation process of insects image classification and semantic annotation used SVM support vector machine method; Finally, the experimental results were offered which showed that the method not only could make up the shortage of the traditional manual image annotation of insects by entomologist, but also able to provide more realtime, accurate semantic annotations on insect images.
Feature extraction Image classification Image annotation
XIE Ming ZHI Min
Computer and Information Engineering College Inner Mongolia Normal University Hohhot, Inner Mongolia, CO 010022, China
国际会议
重庆
英文
428-431
2011-01-21（万方平台首次上网日期，不代表论文的发表时间）