Keyframes detection and analysis in vocal folds recordings using hierarchical motion techniques and texture information

Abstract

This paper presents the study of vocal videostroboscopic videos to detect morphological pathologies using a combination of motion information and segmentation. The motion permits us to obtain the keyframes of total (or minimum) closure and maximum opening and to have the initialization for the segmentation process. The segmentation is made analyzing the image textures applying Gabor filtering. After this stage, where the pathology is (if it exists) is delimited using derivative techniques and interpolation. Finally, the pathology is localized and we can extract some features of each fold.

Laryngeal Image Processing of Vocal Folds Motion

“…A similar strategy has been implemented in [72], which apply an algorithm based on differences between consecutive frames. Other works [11,97,98] use motion estimation techniques to compute the ROI based on the fact that the region with the most salient motion features corresponds to the vocal folds.…”