IMPACT OF HAIR ARTIFACT REMOVAL ON SKIN LESION SEGMENTATION IN DERMOSCOPIC IMAGES

This study evaluates the effect of hair artifact removal on dermoscopic image segmentation. Original and hair -removed images were segmented under identical experimental conditions using the same SK-U-Net model. The results showed improvements in Dice, sensitivity, and overall segmentation accuracy, indicating that hair removal can enhance lesion boundary detection and segmentation reliability.

Hali tarjima qilinmagan