机构地区: 南京财经大学信息工程学院
出 处: 《现代图书情报技术》 2011年第9期41-45,共5页
摘 要: 提出一种基于组合特征的在线商标图像检索系统设计方法,该方法利用颜色直方图匹配、二维傅立叶形状描述子,及二维傅立叶变换形状描述方法与Zemike矩相结合的组合形状描述子,从粗到细、多层次提取商标图像形状特征,用欧氏距离度量形状相似性。用标准MPEG-7的CE2B图像库和1000多张商标图像库进行检索实验,并与现有方法进行比较,表现出其较好的检索性能。 This paper introduces a method for trademark retrieval system using combination features. In this method, a color histogram is first generated for color description, 2D Fourier transform and Zernike moment are then combined to apply to shape features extraction. This method is evaluated on MPEG -7 CE2B and 1 000 trademark images database and is compared with the existing methods. The experimental results show that the performance of this method is acceptable.
领 域: [文化科学]