Retrieving of remote sensing images based on content-sensitive Bayesian networks and direction relations
HU Yuxi1, 2, 3, LI Yikun2, 3, YANG Ping2, 3
1. Xi’an Mapping and Printing Company of ARSC, Xi’an 710054, China; 2. Faculty of Geomatics, Lanzhou Jiaotong University, Lanzhou 730070, China; 3. Gansu Provincial Engineering Laboratory for National Geographic State Monitoring, Lanzhou 730070, China
摘要如何在遥感图像数据库中快速、准确地找出目标图像,是检索系统的核心所在。基于对上下文语境敏感的贝叶斯网络(content-sensitive Bayesian network,CSBN),建立了含有方向关系的检索模型,并根据城市区域的特点,提出了适合城市区域检索的方法。首先,通过贝叶斯网络对图像进行检索; 然后,依据图像的平均高频信号强度(average high frequency signal strength,AHFSS)对候选图像进行筛选; 最后,得到含有城市区域这一高级语义特征的最终检索结果。为了确定图像内部的方向关系,采用东北、西北、东南和西南4个区域的方向描述图像的8种方向关系,有效降低了算法的时间复杂度。实验结果表明,该方法可有效地描述图像的场景语义,并具有较高的查准率和检索效率,可满足用户的需求。
Abstract：Retrieving the required remote sensing images effectively and accurately is the kernel of a remote sensing retrieval system. In this paper, the authors proposed a direction based retrieval model based on context-sensitive Bayesian network(CSBN). In addition, an approach was also proposed that is suitable to retrieving urban area images according to the characteristics of urban areas. Initially, the proposed approach retrieved the candidate images based on CSBN. Then, the proposed approach obtained the final retrieval result containing the high level semantic concept “urban area” according to the average high frequency signal strength(AHFSS)of the candidate images. In order to make sure the direction relationships inside the image, the authors used the four directions of northeast, northwest, southeast and southwest to describe eight kinds of directions, which effectively reduced the time complexity of the algorithm. The experimental results show that the proposed approach can effectively describe the semantic concepts of the stored remote sensing images, and thus has higher retrieval precision and efficiency than the original context-sensitive Bayesian network based approach, thus proving that the proposed approach can meet the users’ requirements.
胡玉玺, 李轶鲲, 杨萍. 基于上下文敏感的贝叶斯网络及方向关系的遥感图像检索[J]. 国土资源遥感, 2017, 29(3): 70-76.
HU Yuxi, LI Yikun, YANG Ping. Retrieving of remote sensing images based on content-sensitive Bayesian networks and direction relations. REMOTE SENSING FOR LAND & RESOURCES, 2017, 29(3): 70-76.
[1] Zhang B.Intelligent remote sensing satellite system[J].Journal of Remote Sensing,2011,15(3):415-431. [2] Li Y K.Semantic-Sensitive Remote Sensing Imagery Retrieval[M].Beijing:China Environmental Science Press,2014:1-7. [3] Wang M,Song T Y.Remote sensing image retrieval by scene semantic matching[J].IEEE Transactions on Geoscience and Remote Sensing,2013,51(5):2874-2886. [4] Datta R,Joshi D,Li J,et al.Image retrieval:Ideas,influences,and trends of the new age[J].ACM Computing Surveys,2008,40(2):Article No.5. [5] Belloulata K,Belallouche L,Belalia A,et al.Region based image retrieval using shape-adaptive DCT[C]//Proceedings of 2014 IEEE China Summit and International Conference on Signal and Information Processing(ChinaSIP),Xi’an,China:IEEE,2014:470-474. [6] Datcu M,Seidel K.Human-centered concepts for exploration and understanding of earth observation images[J].IEEE Transactions on Geoscience and Remote Sensing,2005,43(3):601-609. [7] 林明泽,李轶鲲,安新磊,等.简单贝叶斯网络的遥感图像检索[J].云南民族大学学报(自然科学版),2010,19(1):67-70. Lin M Z,Li Y K,An X L,et al.Remote sensing image retrieval based on the simple Bayesian network[J].Journal of Yunnan University of Nationalities(Natural Sciences Edition),2010,19(1):67-70. [8] Li Y K,Bretschneider T R.Semantic-sensitive satellite image retrieval[J].IEEE Transactions on Geoscience and Remote Sensing,2007,45(4):853-860. [9] Li Y K,Yang S W,Liu T,et al.Comparative assessment of semantic-sensitive satellite image retrieval:Simple and context-sensitive Bayesian networks[J].International Journal of Geographical Information Science,2012,26(2):247-263. [10] 李轶鲲,闫浩文,孙建国.分步式卫星图像检索[J].测绘科学,2009,34(6):53-55. Li Y K,Yan H W,Sun J G.Stepwise satellite image retrieval[J].Science of Surveying and Mapping,2009,34(6):53-55.
2017, Vol. 29 
