fitushar / skin-lesion-segmentation-using-grabcut

Skin lesion segmentation is one of the first steps towards automatic Computer-Aided Diagnosis of skin cancer. Vast variety in the appearance of the skin lesion makes this task very challenging. The contribution of this paper is to apply a power foreground extraction technique called GrabCut for automatic skin lesion segmentation in HSV color space with minimal human interaction. Preprocessing was performed for removing the outer black border. Jaccard Index was measured to evaluate the performance of the segmentation method. On average, 0.71 Jaccard Index was achieved on 1000 images from ISIC challenge 2017 Training Dataset.

Automatic Skin Lesion Segmentation Using GrabCut in HSV Color Space

Skin lesion segmentation is one of the first steps towards automatic Computer-Aided Diagnosis of skin cancer. Vast variety in the appearance of the skin lesion makes this task very challenging. The contribution of this paper is to apply a power foreground extraction technique called GrabCut for automatic skin lesion segmentation in HSV color space with minimal human interaction. Preprocessing was performed for removing the outer black border. Jaccard Index was measured to evaluate the performance of the segmentation method. On average, 0.71 Jaccard Index was achieved on 1000 images from ISIC challenge 2017 Training Dataset. In this work, a framework was proposed for skin lesion segmentation based on automatic GrabCut segmentation. Auto Extracting mask and rectangle initialization strategies was shown for making the segmentation algorithm automatic and generic. The algorithm achieved over 0.71 average Jaccard index for 1000 test images. Future work will be focused on exploring different color channels to improve the performance.

Qualitative and Quantitative results of segmentation