A Novel Medical Image Texture Retrieval Method by Dual-Tree Complex Wavelet Transform

摘要：

As for the problems of shifting sensitivity and poor directionality in discrete wavelet transform (DWT), a novel texture retrieval method of medical images by dualtree complex wavelet transform (DT-CWT) is proposed. In this method, generalized Gaussian statistic model (GCD) is first constructed according to shifting invariability and multi directional selection of DT-CWT coefficients. Then image features are extracted with the above model. Texture retrieval is carried out with improved the log-likelihood (ILL) similarity measurement algorithm. Experimental results show that the veracity of the proposed method is better than the traditional Gabor and DWT algorithm and up to 82.8％. So it is valuable for aid-diagnosis in the future.

关键词： lmage retrieval texture characteristic DT-CWT ILL similarity measurement DWT

作者: Nuan Feng Da-ke Chen

作者单位: Polytechnic School Shenyang Ligong University, Fushun, China Information Science School,Jilin University of Technology, Changchun, China

会议类型: 国际会议

会议名称: 2010 International Conference on Image Analysis and Signal Processing(2010 图像分析与信号处理国际会议 IASP 10)

会议地点: 厦门

会议语种:英文

页码: 97-100

在线出版日期: 2010-04-12（万方平台首次上网日期，不代表论文的发表时间）