Related Experiment Videos
Detecting the edge of the tongue: a tutorial
1Haskins Laboratories, New Haven, CT 06511, USA. iskarous@haskins.yale.edu
Clinical Linguistics & Phonetics
|October 7, 2005
Summary
This tutorial introduces Active Contours for robust tongue edge detection in ultrasound images for speech science. This method effectively handles noisy data and is integrated into current software.
Area of Science:
- Speech Science
- Phonetics
- Medical Imaging
Background:
- Ultrasound imaging is crucial for speech science and phonetics research.
- Accurate tongue edge detection is vital for analyzing speech production.
- Existing methods struggle with speckle noise in ultrasound data.
Purpose of the Study:
- To provide a tutorial on Active Contours for tongue edge detection in ultrasound scans.
- To introduce a robust method for speech science and phonetics researchers.
- To highlight advancements in image processing for future applications.
Main Methods:
- Active Contours (snakes) algorithm for edge detection.
- Assumption of smooth curves in image data.
- Implementation in existing software packages for ultrasound image analysis.
Main Results:
- Active Contours demonstrate robustness against speckle noise.
- The method facilitates accurate tongue edge detection.
- Successful integration into current speech analysis software.
Conclusions:
- Active Contours offer a reliable approach for tongue edge detection.
- Future research may incorporate scale-space and Kalman filters.
- Advancements in image processing will enhance speech production analysis.