Skin Lesion Segmentation Using an Ensemble of Different Image Processing Methods

In recent times, there has been a huge increase in the average number of cases skin cancer per year, which sometimes become life threatening for humans. Early detection various diseases through automated techniques plays crucial role. However, presence numerous artefacts makes this task challenging. Dermoscopic images exhibit variations, including hair artefacts, markers, and ill-defined boundaries. These make automatic analysis lesion quite difficult task. To address these issues, it is essential to have an accurate efficient method will delineate from rest image. Unfortunately, due several types no such thresholding that can provide sufficient segmentation result every type lesion. overcome limitation, ensemble-based proposed selects optimal based on objective function. A group state-of-the-art different methods as Otsu, Kapur, Harris hawk, grey level are used. The obtained superior results (dice score = 0.89 with p-value ≤ 0.05) compared other (Otsu 0.79, Kapur 0.80, hawk 0.60, 0.69, active contour model 0.72). experiments conducted study utilize ISIC 2016 dataset, publicly available specifically designed skin-related research. Accurate help early many diseases.