| Currently, cardiovascular disease has become one of the major causes of human death. In order to the image of the heart to diagnose heart disease, medical imaging technology has become an important clinical tool. Ultrasound imaging diagnostic techniques as a result of its rapid and painless to the patient has become one of the main methods. The speckle tracking imaging technology(STI) is about two-dimensional ultrasound images, based on the best pattern-matching technology used to identify myocardial speckle tracking the movement and tracking of each frame in the image on the location of changes in mapping out the different images in the same frame trajectory of the myocardial location.This thesis is based on the two-dimensional cardiac ultrasound images of the movement sequence, by tracking its short-axis image speckle information, to analyze myocardial strain. The main work is divided into two parts:(1) Extracting feature points in cardiac ultrasound images:According to the characteristics of ultrasound images, this article introduced conception of ultrasound speckle and four pivotal characteristic detection technologies:Forstner, SUSAN, Harris and SIFT.(2) Feature points matching and tracking:Speckle tracking and motion analysis. This article take the SIFT feature matching algorithm implementation of ultrasonic heart image feature point extraction and matching tracking. Firstly, by means of scale-space extrema detection, feature point localization, orientation assignment, feature point descriptor, the feature extraction in cardiac ultrasound images has been achieved. Each feature point has three information:location, scale, direction.(3) Motion analysis for feature points:Registration to be spotted through the matching of corresponding points, which can be time-varying speckle images in different changes of position, displacement can be calculated. According to elasticity of the displacement and strain in the relationship, we can get the general movements of myocardial tissue in different periods. |