Fast K-Means Technique for Segmentation of Dermoscopic Images

Skin cancer remains one of the hazardous cancers diagnoses each year and accounts for more ‎than 50%. When detected at its early stage, standard and cost-effective therapies can treat it ‎successfully - classifying skin sores carefully is key to automating early end systems and ‎mitigating risk factors. Melanoma can provide both shape features and areas of desire for ‎surface examination. Unfortunately, its prevalence can be both unpredictable and lethal, ‎leading to multiple times more passes than all other skin cancers combined. The proposed ‎method demonstrates about the Fast K-Means clustering technique. It can be applied on the ‎dermoscopic skin lesion images which can convert into L*u*v. This approach is then ‎compared with the outcomes of K-Means and Fuzzy C-Means. Evaluation has carried outs ‎includes number of pixels and time complexity analysis. Crystal clear, the proposed method is ‎obviously segmented the lesion and display any visible lesions or spots on them‎.