摘要
针对目前搜索引擎图像检索技术存在语义鸿沟及搜索结果需进一步优化等问题,提出一种基于搜索引擎检索结果的图像过滤方法,以提高图像检索的查准率。以西安为例对其8个著名景点和5种著名饮食的检索结果进行了优化,结果表明:优化后查准率提高了12.7%,验证了该方法的有效性。该方法对展现城市的视觉文化形象具有重要意义。
The images retrieved in internet using the existing search engines need to be refined due to the disadvantages of image retrieval techniques when they are used for the display of urban visual culture image. For this reason,an image filtering method based on the retrieval results of search engine is proposed in order to improve the accuracy of image search. The retrieval results of 8 famous landmarks and 5 famous foods in Xi'an City are refined using this image filtering method,and it is shown that the accuracy of image search is increased by 12. 7% compared with original retrieved images,which verifies the effectiveness of this method. The image filtering method is of important significance to the display of urban visual culture image.
出处
《西安石油大学学报（自然科学版）》
CAS
北大核心
2014年第5期107-110,10,共4页
Journal of Xi’an Shiyou University（Natural Science Edition）
基金
陕西省科技厅计划项目"基于计算机图像检索技术对互联网中西安城市视觉形象研究"(编号:2012K06-11)
