Survey of Techniques of High Level Semantic Based Image Retrieval
Survey of Techniques of High Level Semantic Based Image Retrieval
复制标题
DOI：
--
复制
发表时间：
2013-01
期刊：
影响因子：
--
通讯作者：
H. Kaur;K. Jyoti
中科院分区：
文献类型：
--
作者：
H. Kaur;K. Jyoti
本文综述了近年来基于内容的图像检索技术的研究成果。多年来，图像被用于各个领域，包括反社会因素，力量，保护财产，建筑工程，装饰和服装贸易，新闻和广告，伤害和疾病的治疗，空间数据管理和分析以及文化遗产研究，以提供高效的服务。但是在过去的几年里，由于互联网的发展，图像的数量大量增加。因此，在庞大而多样的图像集合中，不同领域的用户面临着检索与用户查询相关的图像的问题。为了解决这个问题，采用了不同的技术。该调查包括许多出版物，描述了该领域的研究方面，如特征提取，特征匹配，语义差距缩小和性能评估的测量。根据不同的应用需求和目前的技术现状，提出了未来的研究方向。
This paper provides a comprehensive survey of therecent technical achievements in the research area of contentbased image retrieval. Images are being used since many years inall spheres including Antisocial Elements, Forces,Protected property, Construction Engineering, Ornamentationsand rag trade, Press and advertising, Treatment of injuries anddiseases, Spatial data management and analysis and Culturalheritage research for efficient services. But in the past few years,number of images has increased in huge amount due to thegrowth of internet. So, in large and varied collection, users ofdifferent domains face a problem of retrieving images relevant tothe user query. To solve this problem, different techniques areadopted.This survey includes many publications to describe theresearch aspects in this area such as feature extraction, featurematching, semantic gap reduction and measurements forperformance evaluation. Furthermore, based on the demand ofdifferent applications and current technology, a few promisingresearch directions are suggested.