Answers are grounded in this paper and cite their sources.
Skin cancer is one of the most common and dangerous cancers around the world. Early detection of skin cancer can greatly reduce this risk. The main purpose of this paper is to propose an automatic diagnosis system for detecting the healthy and cancerous area. The method includes four main stages including image contrast enhancement, image segmentation, feature extraction and selection, and images classification. Here, 20 different features are employed and the best of them are selected based on a newly introduced optimization algorithm. In the following, this algorithm is also applied to a support vector machine for increasing the classification accuracy. The proposed optimization algorithm is a new improved version of the fluid search optimization algorithm based on chaos theory which increases the system convergence speed. The method is compared with four State of the art methods and shows the superiority of the proposed method toward the analyzed methods.
No takes yet. Share an insight, caveat, or question.
Li et al. (2019) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: