CAJ | 학술논문

传统的基于内容的图像检索技术CBIR系统需要依据图像的可视特征或复合可视特征信息，通过复杂的数学运算进行匹配，面对海量的图像信息上述系统的运算时间将呈线性增长，从而导致效率低下。随着粒度计算理论的完善和成熟，该文建立起基于粒度计算的SSDA图像检索模型，通过控制粒度的大小，提出分层的搜索策略，减少SSDA算法匹配搜索时间，提高图像匹配的效率。
전통적기우내용적도상검색기술CBIR계통수요의거도상적가시특정혹복합가시특정신식，통과복잡적수학운산진행필배，면대해량적도상신식상술계통적운산시간장정선성증장，종이도치효솔저하。수착립도계산이론적완선화성숙，해문건립기기우립도계산적SSDA도상검색모형，통과공제립도적대소，제출분층적수색책략，감소SSDA산법필배수색시간，제고도상필배적효솔。
The traditional content-based image retrieval technology in CBIR system needs based on the visual feature of image or complex visual feature information,and the matching is performed by a complex mathematical operation. The computing time will be increase by linearly and become inefficiencies, when the Image information is massive. Now computing theory is im-proved maturity and perfection,this paper establish a SSDA image retrieval model based on granular computing. By controlling the granular size, the hierarchical search strategy is actualized. It reduce the searching time of SSDA algorithm and improve the ef-ficiency of image matching.