A Visualization System Of Speech Process Based On Ultrasonic And MRI Image Fusion

Economic and cultural exchanges between countries and regions have become increasingly frequent.More and more people need to communicate with foreigners face to face.However,the problem of second language pronunciation inaccuracy is common,which has a great influence on people's speech communication.The traditional teaching of pronunciation adopts the teaching method of following reading imitation,which requires students to find their own problems through auditory feedback,and then make repeated attempts to find the correct pronunciation method.It takes a lot of time and effort.Pronunciation process visualization technology shows pronunciation organs in the process of movement in the form of image display,providing people visual feedback.People can find problems in learning pronunciation and find the correct way of pronunciation more quickly.Visualization of the speech process can also help people with hearing impairment correct their pronunciation.The speech visualization system proposed in this paper is based on the tongue movement image collected by the ultrasonic instrument.A helmet worn by the speaker is used to fix the ultrasonic probe during ultrasonic image acquisition.Ultrasonography and magnetic resonance imaging(MRI)were collected for image annotation processing.Contour curve showing the tongue's surface in the image is extracted.Characteristics based on the curvature of the curve is used as the basis of curve matching.The fusion of ultrasonic and MRI images is realized by matching the contour curve of the tongue surface in the ultrasonic and MRI images.An intuitive tongue movement animation is generated.