基于JPEG的医学图象特征提取  

JPEG Based Medical Image Feature Retrieval

作　　者：苏玉北[1] 黄天锡[1] 廖孟扬[1] 姚永刚[1] 

机构地区：[1]武汉大学电子信息学院医学图象信息及技术研究所

出　　处：《中国图象图形学报（A辑）》1999年第7期583-589,共7页Journal of Image and Graphics

基　　金：国家教委博士点基金

摘　　要：针对医学图象数据库中大量存在的ＪＰＥＧ压缩图象，提出了一种新颖的特征提取算法。该算法提取的特征既包含空域局部信息又包含频域局部信息，全面表达了图象的特征，使得基于内容查询的精度提高；压缩数据只需霍夫曼解码而不必恢复为原始图象即可提取特征，算法耗费时间大为减少；在ＪＰＥＧ压缩图象多分辨率分析的基础上提出了由“粗”到“精”的查询。In this paper, a new image feature retrieval method is proposed for large number of JPEG format medical images.The retrieved result, including both space domain and frequency domain information, expresses image feature perfectly and makes content based query more accuracy. The retrieving can be done just after Huffman decoding but needn't restore the compressed image data to its original state,thus the time consumed by the algorithml is greatly decreased. The method that the query can be done from “rough” to “fine” according to JPEG multi resolution analysis is also presented.

关 键 词：JPEG 特征提取 医学图象 QBIC DCT 多分辨率分析 

分 类 号：R445[医药卫生—影像医学与核医学] TN919.8[医药卫生—诊断学]