Multi-Class Skin Diseases Classification with Color and Texture Features Using Convolution Neural Network

Skin diseases are very common in our daily lives. In the present era, the prevalence of skin diseases is significant, resulting in it being a prevalent health concern. Due to the similarities in the appearance of different skin conditions, automatically classifying them using lesion images poses a challenging task.A small circular or randomly shaped spot on the patient's skin may be identified as indicative of a skin disease. Under certain circumstances, this illness can represent a serious threat when it transforms into skin cancer. Multi-Class Skin Diseases Classification with Colour and Texture Features Using Convolution Neural Network is discussed in this paper. The study investigated several deep learning-based methods for extracting characteristics from various skin cancer images, which are then used with machine learning classifiers to identify the type of skin illness. In the biomedical fields, machine learning algorithms are also widely used for segmentation and diagnosis. To alleviate the burden and assist patients in the early evaluation of a skin lesion, Computer-Aided Diagnosis (CAD) systems have been developed. Therefore, choosing appropriate Machine Learning (ML) algorithms and feature extraction techniques is essential to achieving high classification accuracy. The study's results indicate that integrating characteristics obtained from a Convolutional Neural Network can enhance the effectiveness of categorizing various skin lesions. The performance of the provided architecture is evaluated based on its sensitivity, accuracy, and specificity.