سال انتشار: ۱۳۸۵

محل انتشار: چهارمین کنفرانس ماشین بینایی و پردازش تصویر

تعداد صفحات: ۷

نویسنده(ها):

Mousavi – Control and Intelligent Processing Centre of Excellence, Faculty of Electrical and Computer Eng., Univ. of Tehran, Tehran, Iran
s Setarehdan – Control and Intelligent Processing Centre of Excellence, Faculty of Electrical and Computer Eng., Univ. of Tehran, Tehran, Iran

چکیده:

This study presents a segmentation method for detection of endocardial boundary in echocardiographic images. The proposed method has two stages. Firstly, endocardial boundary is detected by a maximum likelihood segmentation method with parzen estimation of the pdf of the intensity of image. Then, a smoothing method is applied to the resulting boundary to have a smooth curve that is typical for cardiac boundary. Usefulness and potentials of the proposed algorithm is demonstrated by successfully applying it to different images.