基于底层特征的图像内容和人为理解的图像语义之间存在“语义鸿沟"的现象，而基于图像底层视觉特征的图像自动标注技术，能够实现从图像的底层特征中提取出高级语义信息的关键字来标注图像，能很好的解决这一难题-Based on the underlying features of image content and image semantic understanding of human existence between the " semantic gap"　phenomenon, and the underlying image-based visual features automatic image annotation technology that can be achieved from the bottom of the image feature advanced semantic information extracted keywords to annotate images, can be a good solution to this problem