Title :
Texture classification and segmentation using simultaneous autoregressive random model
Author :
Liao, Mengyang ; Qin, Jiamei ; Tan, Yanni
Author_Institution :
Dept. of Radio Inf. Eng., Wuhan Univ., China
Abstract :
The simultaneous autoregressive (SAR) model is used to describe texture. The authors also propose using the least-squares method to estimate six SAR parameters. Based on the SAR model and the parameter estimation method, experiments have been done to classify and segment images of various natural textures and human B-scan images. Excellent results have been obtained
Keywords :
biomedical ultrasonics; image recognition; image segmentation; image texture; least squares approximations; medical image processing; parameter estimation; human B-scan images; least-squares method; parameter estimation; simultaneous autoregressive random model; texture classification; texture segmentation; Gaussian noise; Image classification; Image edge detection; Image segmentation; Liver; Maximum likelihood estimation; Parameter estimation; Pixel; Random variables; Statistics;
Conference_Titel :
Computer-Based Medical Systems, 1992. Proceedings., Fifth Annual IEEE Symposium on
Conference_Location :
Durham, NC
Print_ISBN :
0-8186-2742-5
DOI :
10.1109/CBMS.1992.244923
