摘要
提出一种将三个特征结合的方法来进行商标图像检索。首先用Canny算子对商标图像的边缘进行提取;其次计算出商标图像边缘的7个Hu不变矩,以及偏心率和质心距离,得到一个9维向量;然后用欧氏距离匹配;最后给出了和其它算法对比的实验结果。
This paper proposes a method of trademark image retrieval based on multi-feature extraction. Firstly, the proposed ap-proach commences with the extraction of edges using the Canny edge detector. Secondly, extracts the Hu moment invariant, eccentrieity and eentroid distance. Thirdly, Euclidean distance is used to measure the similarity between the query and database images. At last the performance of the proposed algorithm is compared with other algorithms.
出处
《计算机与现代化》
2009年第10期82-85,共4页
Computer and Modernization
关键词
商标图像检索
特征提取
特征匹配
trademark image retrieval
shape feature extraction
shape feature matching
