Text this: An automatic laryngoscopic image segmentation system based on SAM prompt engineering: from glottis annotation to vocal fold segmentation