CAJ | 학술논문

图像标签的自动标注技术是当前信息检索领域的热点问题。图像标注本质上是一个机器学习问题，该文对基于超像素的CRF图像分类标注和基于纠错输出编码的图像分类标注两类不同标签标注方法进行对比，阐述其优缺点，指出了实现低复杂度的图像分类标签标注方法的途径。
도상표첨적자동표주기술시당전신식검색영역적열점문제。도상표주본질상시일개궤기학습문제，해문대기우초상소적CRF도상분류표주화기우규착수출편마적도상분류표주량류불동표첨표주방법진행대비，천술기우결점，지출료실현저복잡도적도상분류표첨표주방법적도경。
The automatic annotation technology of image labeling is a hot issue in the field of information retrieval currently. Im-age annotation is a machine learning problem essentially. In this paper, two different annotation technologies based on the com-parative analysis, in which one is CRF image classification method based on supper-pixel and the other one is image classification labeling annotation based on correcting output coding which achieves low complexity.